Add a GPU cloud
GPU Economy lists any GPU cloud that publishes a public rate card. A listing is free and cannot be bought: providers are found by us or suggested by anyone, and every price on the site links to the provider’s own page.
What a provider needs to be listed
- A public pricing page, reachable without an account, a login or a sales call. A price we cannot link to is not published, whoever sends it.
- An hourly price per GPU, or per instance with its GPU count. We divide a node price by its GPU count ourselves and say so next to the figure.
- A stated price. “Contact sales” gives us nothing to link to. An advertised “from” price is shown tagged as such and kept out of the Price Index, and we may wait for a full rate card before listing a provider that publishes only that.
- Prices readable without a browser: in the page’s HTML or a public API, JSON being best. Since September 26, 2026 a page that draws its prices with JavaScript can be rendered too, with our own user agent; a bot check, a captcha or a refusal (403, 429) ends the read, and is never worked around.
- A GPU we track: AMD Instinct MI300X, AMD Instinct MI325X, AMD Instinct MI350X, AMD Instinct MI355X, Intel Gaudi 2, NVIDIA A10, NVIDIA A100 40GB, NVIDIA A100 80GB, NVIDIA A10G, NVIDIA A2, NVIDIA A30, NVIDIA A40, NVIDIA B200, NVIDIA B300, NVIDIA GB10 (DGX Spark), NVIDIA GB200 NVL72, NVIDIA GB300 NVL72, NVIDIA GeForce GTX 1070 Ti, NVIDIA GeForce GTX 1080, NVIDIA GeForce GTX 1080 Ti, NVIDIA GeForce GTX 1660 SUPER, NVIDIA GeForce GTX 1660 Ti, NVIDIA GeForce RTX 2060 12GB, NVIDIA GeForce RTX 2080 Ti, NVIDIA GeForce RTX 3060, NVIDIA GeForce RTX 3060 Ti, NVIDIA GeForce RTX 3070, NVIDIA GeForce RTX 3070 Ti, NVIDIA GeForce RTX 3080, NVIDIA GeForce RTX 3080 Ti, NVIDIA GeForce RTX 3090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 4060, NVIDIA GeForce RTX 4060 Ti, NVIDIA GeForce RTX 4070, NVIDIA GeForce RTX 4070 SUPER, NVIDIA GeForce RTX 4070 Ti, NVIDIA GeForce RTX 4070 Ti SUPER, NVIDIA GeForce RTX 4080, NVIDIA GeForce RTX 4080 SUPER, NVIDIA GeForce RTX 5060, NVIDIA GeForce RTX 5060 Ti, NVIDIA GeForce RTX 5070, NVIDIA GeForce RTX 5070 Ti, NVIDIA GeForce RTX 5080, NVIDIA GH200, NVIDIA H100 80GB (form factor not stated), NVIDIA H100 NVL, NVIDIA H100 PCIe, NVIDIA H100 SXM, NVIDIA H200, NVIDIA H200 NVL, NVIDIA L4, NVIDIA L40, NVIDIA L40S, NVIDIA Quadro P4000, NVIDIA Quadro RTX 4000, NVIDIA Quadro RTX 6000, NVIDIA RTX 2000 Ada, NVIDIA RTX 4000 Ada, NVIDIA RTX 4000 SFF Ada, NVIDIA RTX 4090, NVIDIA RTX 4500 Ada Generation, NVIDIA RTX 5000 Ada Generation, NVIDIA RTX 5090, NVIDIA RTX 5880 Ada, NVIDIA RTX 6000 Ada, NVIDIA RTX A2000, NVIDIA RTX A4000, NVIDIA RTX A5000, NVIDIA RTX A6000, NVIDIA RTX PRO 4500 Blackwell, NVIDIA RTX PRO 5000 Blackwell, NVIDIA RTX PRO 6000 Blackwell, NVIDIA RTX PRO 6000 Blackwell Max-Q, NVIDIA T4, NVIDIA Tesla P100 12GB, NVIDIA TITAN RTX, NVIDIA TITAN V, NVIDIA TITAN Xp, NVIDIA V100, NVIDIA V100 32GB. Other models are recorded as unmapped and added deliberately, never guessed.
- Any currency. A price listed in another currency is shown as listed, next to its USD conversion at the dated official rate: the ECB’s, or the Bank of Russia’s for RUB and VND, which the ECB does not publish. Only USD listings enter the Price Index.
Suggest a provider
Add your GPU cloudPrices come from your public pricing page, never from this form. Listing is free and can’t be bought. How providers are listed, and who is not, and why
Or email data@gpueconomy.com with the provider’s name, the address of its public pricing page, the API documentation if prices come from an API, and the GPU models it rents. You do not have to work for the provider.
What happens next: we read the page, run our parser against the live source without writing anything, and check the numbers against other providers for the same card. If the page meets the conditions above, the provider appears with every price linked to that page; if it does not, it is not listed yet. We do not promise a date.
What the form keeps
Your email, the link you sent, the page the form was on and the time — no IP address and no browser details. The email is used only about your listing and never shared; to have a submission deleted, write to the address above.
A price on this site is wrong?
Email data@gpueconomy.com with the page on this site and the provider’s page that shows the correct price. Every price here names where it was read and when; a correction is checked against that page, and a mistake is fixed.
What we never do
- Sell a listing, a position in a table or inclusion in the Price Index. Price tables are sorted by price.
- Publish a price sent to us by email, a quote, or a negotiated rate.
- Let a referral programme change which prices a table shows or in what order — see the link policy.
Who we read
As of , we read the prices of 121 providers for 82 GPU models: airgpu, Airon, Akamai Cloud (Linode), Amazon Web Services, Atlantic.Net, Atlas Cloud, Bamboozle Web Services, Barrack AI, Baseten, Beam, Bentaus, Bitdeer AI, Canopy Wave, Cato Digital, centron, Cerebrium, Civo, Clore.ai, Cloud.ru Evolution, CloudGPU, CloudRift, CoreWeave, Crusoe, Cudo Compute, Dataoorts, Daytona, Deep Infra, Denvr Dataworks, E2E Networks, EcoHash, Edgevana, Elice Cloud, EmpirioLabs, Enverge, fal, FarmGPU, Fireworks AI, FlexAI, FluidCore, FPT AI Factory, Gcore, Geodd, GMI Cloud, Google Cloud, GPU.ai, GPUwerk, Helios Cloud, Hetzner, HexGrid, Hikube, Hivenet, Hostinger, Hot Aisle, HPC-AI.COM, Hugging Face, Hyperbolic, Hyperstack, immers.cloud, Inferless, Infomaniak, Intelion Cloud, IONOS Cloud, Jarvislabs, Koyeb, Krutrim Cloud, Lambda, Leafcloud, Lightning AI, Limestone Networks, Lium, Lyceum, Massed Compute, Meshive, Microsoft Azure, MimicPC, Modal, MWS Cloud Platform, Nebius, Nebula Block, Neysa, Northflank, Nova Cloud, Octa.space, OpenRelay, Oracle Cloud (OCI), OVHcloud, packet.ai, Paperspace, Phala Cloud, QuantaCloud, Replicate, RunComfy, RunPod, Runware, Salad, Scaleway, Seeweb, Selectel, Servercore, Sesterce, STACKIT, STN, SwissGPU, SynpixCloud, Taiga Cloud (Northern Data Group), Tata Communications, TensorCloud, TensorPool, TensorWave, TheAI Cloud, Thunder Compute, Together AI, Trooper.AI, Vagon, Vast.ai, Verda, VESSL AI, Vultr, Yandex Cloud, Yotta Labs, YTL AI Cloud.
Considered and not listed
Every GPU cloud, price source and single quote we looked at and do not list, with the reason and the day we last checked, ordered by reason and then by name. No figure here is listed as a price: where a reason needs a figure to say why - a rate far from every other source, a floor withheld as too thin - it is quoted as evidence, with the day it was read, and enters no table, median or index on this site (until 2026-10-01 this line said no price was shown here at all, which these quotes contradicted). Where the reason is that we cannot read the prices, the limit is ours, not the provider’s.
| Provider or source | Reason | What we saw | Checked | Look again |
|---|---|---|---|---|
| Aethir | No public rate to link to (HTTP 200) | A decentralised GPU network selling enterprise capacity through a Google Form: the home page prints no rate and links no pricing page, and its sitemap lists only blog posts and product pages, none with a GPU-hour figure. | ||
| Aircloud | No public rate to link to (HTTP 200) | Its home page sells private B300/B200 clusters of 32-128 nodes and NVL72 racks "reserved for 48 or 60 months"; the rate table ("Reserved rates per GPU-hour on a 60-month term, by platform") prints only starting figures - "HGX B300 ... from US$4.25 / GPU-hr", "B200 from US$4.00" - and says "The rate is quoted per deal. Longer terms and more prepayment lower it." No on-demand rate and no fixed rate for a term we publish. | ||
| Alta Technologies retail source | No public rate to link to | A reseller that sells datacenter GPUs by quote only, with no public price to link to. | — | |
| Beyond.pl | No public rate to link to (HTTP 200) | Its GPUaaS table says "Contact sales" in every cell but one: a DGX B200 8-GPU weekly rate marked "$5.10/h/GPU*" with the asterisk's own conditions. GB300 is named with no figure. One starred weekly cell is not a published hourly rate card. | ||
| CambridgeNexus | No public rate to link to (HTTP 200) | No pricing page (/pricing 404). The home page names "GB300 NVL72 Racks" and asks the reader to "Request Enterprise Access" - no rate anywhere. | ||
| Cherry Servers | No public rate to link to (HTTP 200) | Its GPU offer lives at /dedicated-gpu-servers (found in sitemap-2.xml). The hourly figures printed there are for servers without a GPU: each card reads "GPU Optional up to 1 GPU" beside a rate such as 0.318 USD/hr, while the GPU itself - A2, A16, A100, T4, Quadro K2200/K4200 are named - is "Configure Server" or "Chat with an Expert" with no rate of its own. | ||
| Clomd | No public rate to link to (HTTP 200) | Its pricing page prints one figure for compute, "H100 GPU pricing starts at $2.39/hour", and says "All pricing is custom based on your minimum spend commitment" with "Get Custom Quote" - a floor for a quote, not a rate a renter is billed. | ||
| CloudSigma | No public rate to link to (HTTP 200) | No list price to read. The pricing page (www.cloudsigma.com/pages/pricing, the home page's link) answers 200 and, served and rendered alike, carries no number beside a currency and no GPU at all: it describes a revenue-share model in which "partners set their own customer pricing" - "Revenue-share, not fixed wholesale list prices". The rate a buyer pays is set by a reseller, not printed by CloudSigma. | ||
| Corvex | No public rate to link to (HTTP 200) | The scout's address www.corvex.io redirects to www.corvex.ai, whose sitemap of 61 URLs names no pricing page. Its per-card pages answer 200 and print their only money figure in the meta description - /product-h200 "starting at $2.15/hour", /product-b200 "starting at $3.19/hr" - while the body a reader sees says "Get a custom quote to reserve your NVIDIA H200s" and the FAQ answers what pricing plans exist with no number at all. A figure in a meta tag is not one the linked page shows, so there is no rate to publish. llms.txt is 404. | ||
| Elastx | No public rate to link to (HTTP 200) | A Swedish OpenStack and Kubernetes operator. Its pricing page carries no figure at all - it describes hourly billing and asks the reader to get in touch for a price. | ||
| ESDS | No public rate to link to (HTTP 200) | The GPU as a Service page (linked from the apex) names L40S, H200, B200, B300 and GB200 with specs and prints no rate: "A solutions architect responds within 24 working hours - scoped to your stack, priced in INR". The only dollar figure on the page is a market size ($4.8 trillion). | ||
| FluidStack | No public rate to link to (HTTP 200) | Its sitemap lists only the home page, jobs, news and legal pages - no priced path exists to probe - and the home page it serves carries no rate, no "per hour" and no pricing section of any kind. Capacity is arranged by talking to the company. | ||
| Global Scale | No public rate to link to (HTTP 200) | Its GPU infrastructure page prints only ranges under "Indicative Pricing - Market rates updated weekly. Prices vary by term length and volume" (e.g. H100 SXM5 "$1.40 - $1.75" per GPU-hour) beside "Get a firm quote"; no single rate a renter is billed. | ||
| GPUaaS | No public rate to link to (HTTP 200) | Its cluster pages name the cards it offers (B300, B200, H200, H100, A100, RTX 5090, RTX 4090, RTX 6000) and print no rate for any of them; the pages lead to a checkout rather than a published price. | ||
| GPUnex | No public rate to link to (HTTP 200) | No GPU-hour rate to read. The rent page (www.gpunex.com/rent/, the home page's link) answers 200 and, served and rendered alike, names H100, A100, L40S and L4 with no number beside a currency: access is by application ("Apply for GPU Access", "applications are typically reviewed within 24-48 hours"). The only dollar figures on the site are the GPU "packages" it sells to investors with daily payouts, which are not a rental price. | ||
| Green AI Cloud | No public rate to link to (HTTP 200) | No pricing page (/pricing 404). A banner reads "B200 ... rates starting at 3,75 USD/GPU/H" and then sends the reader to "Contact us" - a starting-from banner, no rate card and no other card priced. | ||
| IBM Cloud | No public rate to link to (HTTP 200) | Rendered GPU product and pricing pages name A100, L40S, MI300X and B300 with no published hourly rate. | ||
| IREN (IREN Cloud) | No public rate to link to (HTTP 200) | iren.com/solutions/gpu-cloud/ai-cloud answers 200 and lists five cards - H100, H200, B200, B300 and GB300 NVL72 - each with its specs and the words "Contact sales" where a rate would be; the page offers "exclusive pricing" when you "deal directly with IREN". Its sitemap carries no pricing path at all. Nothing to read until a rate is printed. | ||
| ITGLOBAL | No public rate to link to (HTTP 200) | The GPU cloud server page (linked from the apex) prints no figure at all, served or rendered: it lists cards and options and ends "To receive a custom quote for cloud server with GPU price, please fill out the form below". The price depends on VMs, configuration and rental period; no hourly rate per card is published. | ||
| KINX IXcloud | No public rate to link to (HTTP 200) | Its service price page (linked from the home page) prints hourly and monthly rates in KRW for CPU instances, but the "GPU instance" row (4-16 vCore, 16-62 GB, NVIDIA Tesla series options) reads "To be negotiated separately" - "The GPU instance service fee is set individually ... please submit your inquiry". No GPU rate is published. | ||
| Nebul | No public rate to link to (HTTP 200) | Its GPU page names B300, L40S and L4 and prints no price, rendered as a visitor's browser draws it (314 KB of rendered HTML, not one currency figure); the page offers "Talk to us" instead. The served HTML had already shown no figure on 2026-09-27. | ||
| Nscale | No public rate to link to (HTTP 200) | Its sitemap has no pricing page at all; the priced paths are per-card pages (/gpu-nodes/nvidia-h100, -h200, -gb200, -gb300-nvl72), and all four are served as static HTML with no figure on them - the call to action is "Reserve GPUs" and "Get in touch". The page even carries a schema.org Offer whose priceSpecification names the currency (USD) and the unit (1 GPU node) and leaves the price out. | ||
| Parasail | No public rate to link to (HTTP 200) | No GPU-hour rate to read. www.parasail.io/pricing (the home page's link) answers 200 and prices inference per token; its dedicated hardware table names B300, B200, H200, H100 SXM, RTX PRO 6000 and RTX 5090 with "Get a quote" in every row and the line "Hourly rates available on request · volume discounts available". Reserved capacity is "negotiated". A quote on request is not a published price. | ||
| puzl.cloud | No public rate to link to (HTTP 200) | An Estonian cloud whose pricing page (puzl.cloud/pricing/, the home page's link) answers 200 and prices only CPU, memory and storage per utilized second ("EUR0.00002 per utilized vCPU-second"); GPUs carry no figure: "If you require resources for a regular usage or a specific hardware configuration, including GPU or TPU, we are happy to offer you an individual long-term contract" via sales. No GPU rate to read. | ||
| Radian Arc | No public rate to link to (HTTP 200) | No published rate and no priced path. The company sells "Sovereign AI infrastructure & GPU platform" and its apex page carries 135 KB with two dollar figures, neither of them a price for compute; robots.txt is served empty and names no sitemap, and sitemap.xml answers 404, so there is no index of pages to find a rate card in. This is the company the aggregators still list as Cloudalize, whose domain now redirects here. | ||
| Runyour AI | No public rate to link to (HTTP 200) | No rate on any public page. The marketing site names the H100 and the A100 and prints no figure in any currency; its console page for GPU Cloud, rendered under our own user agent, shows a shell whose only offers are "Quote Inquiry" and an invitation to rent out your own computer. The machine book, and any price in it, is behind a sign-in. | ||
| Sakura Internet | No public rate to link to (HTTP 200) | Its GPU line-up prices services, not named cards: the container product DOK at "0.016 yen/second (57.6 yen/hour) and up" and the VM product VRT at "481 yen/hour and up", tax included, each for a service whose card is one of V100 or H100; the bare-metal product PHY is "please contact us". Floors for a service are not a rate for a card. Re-probed 2026-09-28; yen is no longer a reason. | ||
| SF Compute | No public rate to link to (HTTP 200) | The pricing path redirects to a login; the home page is a spot and reserved market with no rate card. | ||
| Sharon AI | No public rate to link to (HTTP 200) | An Australian neocloud with announced capacity contracts in the billions and not one money figure on its own site: 594 KB served and 205,131 characters of rendered text, with zero numbers standing beside a currency symbol or an hour unit in either. It sells capacity on take-or-pay contracts, and the page's own calls to action ask for contact rather than naming a rate. Nothing to read, nothing to publish. | ||
| Timeweb Cloud | No public rate to link to (HTTP 200) | Its GPU page names the cards it rents (H200, H100, A100, A5000, A4000, A2000, A30, A2, T4) with benchmark scores, and prints no hourly rate for any of them in the served HTML: the only rouble figures are a minimum advance payment and grant sums. Re-probed 2026-09-28, when the currency stopped being a reason. | ||
| Voltage Park | No public rate to link to | The pricing page gives only a “starts at” figure for H100s, and its on-demand block reads “Contact for pricing”; the rate card itself sits in the logged-in dashboard. | ||
| WhiteFiber | No public rate to link to (HTTP 200) | Its real priced path (/gpu-pricing, found in its own sitemap after a guessed /pricing returned 404) is static HTML carrying exactly one sentence about price - "Contact us to discuss pricing based on your specifications" - and not a single figure. | ||
| zCLOUD (Zettabyte) | No public rate to link to (HTTP 200) | Behind a "Get Early Access" form. The only figures are "from $2.51 / GPU-hour" on demand and "from $2.20" reserved, with no card named at all, so no figure can be attached to a GPU. | ||
| Zenlayer | No public rate to link to (HTTP 200) | No price path answers: /pricing/ 404 and the GPU product page 404. The home page carries no rate. | ||
| 1Legion | On hold: our decision (HTTP 200) | Reserved-only 8-GPU servers, and the page prints two different 12-month rates for the same card (RTX 5090 from $1.00 in the table, $0.59 in the comparison; H100 SXM $2.40 vs $1.99). | ||
| AIEEV | On hold: our decision (HTTP 200) | Rendered pricing page (Wix) shows no GPU-hour price. | ||
| Alibaba Cloud | On hold: our decision (HTTP 200) | Rendered ECS price list shows no GPU instance rate without choosing region and family by hand. | ||
| Anyscale | On hold: our decision (HTTP 200) | Its pricing page answers 200 and prices compute in Anyscale Credits, not a currency: "NVIDIA T4 AC 0.5682 /hr", "NVIDIA L4 AC 0.9542 /hr", "NVIDIA A10G AC 1.3635 /hr", "NVIDIA A100 AC 4.9591 /hr" (the A100's memory not named), H/B/GB "Contact Us", under a Hosted / BYOC switch that does not say which deployment the figures are for. The page nowhere states what one credit costs in dollars, so no rate can be published. | ||
| AtmosCompute | On hold: our decision (HTTP 200) | Per-card pages list shifting configurations that look like resold marketplace offers (0 vCPU rows, 45 GB A40, 20 GB RTX 3090); what machine a row is cannot be verified on the page. The page read on 2026-09-26, atmoscompute.com/prices/gpu/a40, answers 404 since (checked 2026-10-03), so the link is the site itself. | ||
| BreezeHost | On hold: our decision (HTTP 200) | One card (RTX 5090), out of stock; the on-demand rate appears only in the page title, the body prints a spot "from" price. | ||
| Cleura | On hold: our decision (HTTP 200) | Its public price list (the JSON its pricing page loads) carries VMs and storage but no GPU; no GPU rate found after rendering. | ||
| Ekasunucu | On hold: our decision (HTTP 200) | Rendered GPU server page names H100 and A100 with no hourly price (monthly servers, TRY by default). | ||
| Exoscale | On hold: our decision | Its price file lists the same figures under EUR, CHF and USD, and the page switches the currency label without changing them, so we cannot tell which currency a GPU rate is in. | ||
| GPUhub | On hold: our decision (HTTP 200) | Rendered page and its own open API disagree about tenfold (RTX PRO 6000 $1.16/hr on the page, $11.02 per GPU-hour in the API); which is the price cannot be told. | ||
| Hinode | On hold: our decision (HTTP 200) | Machines priced by VRAM and RAM "from" (L4 24 GB from $1.75/hr), the cheapest being fractions of a GPU; no whole-GPU rate to compare. | ||
| HyperAI | On hold: our decision (HTTP 200) | GPU hours come with a paid membership (the free plan has a GPU quota of 0 hours), so the hourly rate on the page is not the whole price of an hour. | ||
| ionstream | On hold: our decision (HTTP 200) | Its solutions page prints one "from" rate per GPU-hour for each card (B200, B300), with no term. The company told us in writing that these rates assume about a quarter of the server paid up front and that the B200 is no longer available. Until the page states the terms a rate is sold on, it cannot be listed as an hourly rental price. | ||
| Latitude.sh | On hold: our decision | Every GPU plan is priced “Starting At”, the cheapest location. Our decision: not listed until plans are priced per location, so that a floor never enters the Price Index as a rate. | ||
| OmegaGradient | On hold: our decision (HTTP 200) | Open marketplace catalogue with one to four offers per card - below the three verified offers a marketplace floor needs. | ||
| Qubrid | On hold: our decision (HTTP 200) | GPU prices sit behind a "GPU Compute" tab that takes a click; rendering shows only per-token model prices, and clicking is not reading. | ||
| QuickPod | On hold: our decision (HTTP 200) | Rendered home page carries no GPU-hour price; offers appear only inside the signed-in app. | ||
| SimplePod | On hold: our decision (HTTP 200) | Two pages of its own print different hourly figures for the same card, and the lower one is not what a buyer can pay. The pricing page (simplepod.ai/pricing, static) lists "GeForce RTX 4090 24 GB $0.30/hour", RTX 5090 $0.45, RTX 6000 Blackwell $0.99; the home page's live table of host offers says RTX 4090 "Docker: from 0.41$/hour", RTX 5090 "from 0.542$". A list price below the cheapest live offer is stale or promotional; the live table is a marketplace "from" (one cheapest host), which we do not publish as a floor. | ||
| SpotGPUs | On hold: our decision (HTTP 200) | Prices far below every other source (A100 80GB $0.29, GH200 $0.79 on demand) and payment in crypto; not taken until the offering can be verified as a rentable machine at that rate. | ||
| Theta EdgeCloud | On hold: our decision (HTTP 200) | The pricing path renders "Page not found"; no GPU rate on the rendered site. | ||
| Arkane Cloud | We cannot read the prices (HTTP 403) | Its pricing page answers the GitHub runners the hourly scrape reads from with HTTP 403 (a 5,397-byte refusal page, 2026-10-24 10:09 UTC), while it served the routine's container 200 with a static table of per-GPU hourly rates. A server refusal to the reader that would read it every hour, so the page is not read; asked once from the runner, not retried. | ||
| CloudPe | We cannot read the prices (HTTP 403) | Its static price page (an hourly rupee rate per GPU machine) reads from a desk, but answered 403 Forbidden to the first hourly read from our GitHub Actions runners (2026-10-07, 16:37 UTC), so no quote was ever stored. A refusal is not worked around; the parser module was withdrawn and is re-added only if a re-probe from the runners is answered. | ||
| ConoHa (GMO) | We cannot read the prices (HTTP 403) | Cloudflare answers its robots.txt itself with an interactive challenge ("Just a moment...", HTTP 403) to a plain client, so not even the crawling rules can be read, let alone a price page. A challenge is a refusal: it was asked once and not retried, and no browser user agent was tried. | ||
| Contabo | We cannot read the prices (HTTP 403) | Its GPU server page answers our own user agent 403 (the en/ path redirects to en-us/ and refuses there). A server's refusal is a refusal we accept: not retried, not worked around. | ||
| Cyfuture AI | We cannot read the prices (HTTP 403) | Its static price page reads from a desk, but answers 403 Forbidden to every hourly read from our GitHub Actions runners (2026-09-28, 20:37 and 21:37 UTC), so no quote was ever stored. A refusal is not worked around; the parser module was withdrawn and is re-added only if a re-probe from the runners is answered. | ||
| HighReso (GPUSOROBAN) | We cannot read the prices (HTTP 200) | Its per-card rates are not text a reader or parser can verify. /price answers 200 and, served and rendered alike, carries one figure in words, "1時間50円から" (from 50 yen an hour), plus links to /compute and /multinode-cluster; the price tables are images (price_compute, price_aispacon, price_remowa). /compute, probed and rendered on 2026-10-08 10 UTC, printed the same "from" sentence only. | ||
| Intel Tiber AI Cloud | We cannot read the prices (HTTP 200) | ai.cloud.intel.com, the address aggregators give for its Gaudi 2 rates, does not resolve (GitHub runner, 2026-10-26 04:08 UTC: name not known). cloud.intel.com answers 200 with "Intel Cloud Services - Pre-release Hardware", a sign-in app: rendered, it prints no GPU or accelerator price and links no pricing page. Prices, if any, are only inside the signed-in console. | ||
| io.net | We cannot read the prices (HTTP 401) | No pricing page (/pricing 404); the hardware price API answers 401 without authentication. | ||
| Micron21 (mCloud) | We cannot read the prices (HTTP 200) | Named by the scout on 2026-10-04 and without a verdict for five runs: /enterprise/mcloud answers 200 with 168 KB and no number beside a currency, so it needed a render, and the render timed out waiting for the page to go idle in four runs. On the fifth the load completed and the renderer refused it - what came back matched its bot-check markers, which are never bypassed. The confirming load timed out again, so the marker was not captured; four timeouts fit a challenge page that never goes idle, but that is a correlation, not a cause. | ||
| NeevCloud | We cannot read the prices (HTTP 403) | Its price page answers 403 to a plain request from our own user agent; bot protection is never worked around, so the page cannot be read. | ||
| Novita AI | We cannot read the prices | GPU instance prices are drawn in the browser from a keyed API; none are in the page HTML. | ||
| NTT (GPUaaS) | We cannot read the prices (HTTP 403) | The home page www.ntt.com answers a plain client with HTTP 403, so not even the link to its GPU-as-a-service page can be read. Asked once, not retried, and no browser user agent was tried. | ||
| Oblivus | We cannot read the prices (HTTP 520) | Its pricing page answers our reader with a 520 error and then refuses every request from the same address for about a minute; reading it would mean getting around that protection, which we do not do. | ||
| Reg.ru (Cloud GPU) | We cannot read the prices (HTTP 200) | The GPU cloud page answers 200 with a 13 KB stub instead of the page: a script computes a value, sets it as the cookie __jhash_ together with the browser's user agent, and reloads (location.reload()). That is a bot check, so it was not passed: a headless render of the page did not settle either (the page kept navigating). No price was seen. | ||
| Scaleway | We cannot read the prices (HTTP 403) | Its GPU price page, read hourly since September, answers our reader 403 with a Bunny Shield bot check ("Establishing a secure connection") instead of the page since 2026-10-27, 10:24 UTC - from GitHub's runners and from our own container alike. A challenge is a refusal and is not passed; the hourly run stops asking, the module stays for a re-probe, and its last rates keep their read time. | ||
| ServerSupply retail source | We cannot read the prices (HTTP 403) | A reseller of datacenter GPUs whose site refuses our reader, so its listings cannot be linked or checked. | — | |
| Tencent Cloud | We cannot read the prices (HTTP 200) | Its site answers our reader with a bot check instead of a page. www.tencentcloud.com and its GPU billing page (document/product/560/8025, which served 29 KB of text on 2026-10-11) both answer 200 with a 987-byte script that sets an "EO_Bot_Ssid" cookie and reloads - a challenge we do not pass. Even before, that page printed no GPU-hour figure. | ||
| TensorDock | We cannot read the prices | Its marketplace API, where machines and their prices were listed, returns an empty host list since TensorDock was acquired by Voltage Park. | ||
| UpCloud | We cannot read the prices (HTTP 403) | robots.txt is served and open, but the pricing page answers a plain client with HTTP 403, so the rates cannot be read. Asked once, not retried, and no browser user agent was tried. | ||
| Utho | We cannot read the prices (HTTP 403) | Cloudflare answers both robots.txt and the pricing page with 403 and a "Just a moment..." bot check. A server refusal, so the page is not read and the check is never passed. | ||
| Cloudalize | Not trading (HTTP 200) | The brand no longer exists to price. cloudalize.com answers 200 but redirects to www.radianarc.com, whose page is titled "Radian Arc | Sovereign AI infrastructure & GPU platform" and names Cloudalize nowhere in its 135 KB; there is no pricing path and no hourly figure on it. The www host is worse than that - it answers a plain TLS client with a handshake failure - so there is nothing to read under this name at any path. Radian Arc itself is a separate candidate and is noted in the queue, not here. | ||
| CloudFerro | Not trading (HTTP 404) | No GPU price page can be found: /pricing/, /en/pricing/ and /pricing-list/ all answer 404, and robots.txt answers 403 HTML. | ||
| ComfyDeploy | Not trading (HTTP 200) | Closed to new customers: app.comfydeploy.com/pricing (a 708-byte shell, prices drawn by script) rendered shows a per-GPU-hour list (T4 $0.6480, L40S $2.1456, H100 $8.4168, B200 $9.3744 and others) under the banner "We are no longer taking new customers. Existing customers: service and support remain unchanged." A rate a new reader cannot buy is not listed. | ||
| CR8DL | Not trading | The domain does not resolve at all: no A record for cr8dl.ai from this container and ENOTFOUND on a second, independent path. There is no page to read at any address, and the name reached us through an aggregator CSV that has evidently gone stale. | ||
| Fly.io | Not trading (HTTP 200) | The pricing page answers 200 and no longer lists any GPU: the GPU product it had is not on it. | ||
| Genesis Cloud | Not trading (HTTP 522) | Reported to be in liquidation, and its site does not load: the origin server is unreachable. | — | |
| Hexabyte | Not trading (HTTP 200) | Swedish host (Umea) with four NVIDIA L4 plans priced by the hour in kronor - l4.gpu1 3.91, l4.gpu2 4.28, l4.gpu3 4.82, l4.dual-gpu 10.69 for two cards, ex VAT - but the service cannot be bought. The table is headed "Prisexempel" (price examples) and the only call to action under every plan, and in the page's hero, is "Ga med i vantelistan": join the waiting list. The figures are drawn by a script; a headless render under our own user agent showed them, so this is not a reading problem. Publishing them would put an hourly rate on the site that nobody can order. | ||
| Kawkav | Not trading (HTTP 504) | The apex links its GPU service ("GPU On-Demand ... Open") at gpu.kawkav.com, and that address answered 504 Gateway Timeout on three reads days apart (2026-10-25, 2026-10-26, 2026-10-30), one read each, not retried. The apex itself prints one figure, "from EUR 0.39 /hr" for "A4000 to A100/H100 by the minute", tied to no card. | ||
| Lepton AI | Not trading (HTTP 200) | The brand no longer has a site of its own: lepton.ai answers 200 and redirects to www.nvidia.com/en-us/data-center/dgx-cloud-lepton/, which prints no price of any kind - no dollar figure and no hourly rate in 315 KB - and has no pricing path. The www host does not resolve. The page it lands on is NVIDIA's marketplace of other providers' GPU capacity, so a rate there would be another operator's quote rather than its own. | ||
| Ori | Not trading (HTTP 200) | The pricing page redirects to radiant.co/ori, which states "Ori has merged with Brookfield-backed Radiant". No price page survives the merge. | ||
| 3NV | Not a GPU-hour (HTTP 200) | An Estonian VM host, not a GPU cloud: its pricing page sells vCPU, RAM and NVMe by the month in euro (3.49 EUR per vCPU, 2.99 EUR per GB of RAM) and the word "GPU" does not appear on it at all. The aggregator that listed it names GPUs only in its own "alternatives" sidebar. | ||
| AceCloud | Not a GPU-hour (HTTP 200) | Prices GPUs by the month only, with no stated hours-per-month divisor. The INR card page for the L40S (Noida) answers 200 with a table whose Hourly column reads "---" in every row; the figures are Monthly, 6-month and 12-month (N.L40S.64 1x GPU, 16 vCPU, 64 GB: ₹83,000 a month). The USD page (Atlanta, read 2026-10-08) is the same: Hourly "---", $1,365 a month. No hourly rate can be published. | ||
| AIME | Not a GPU-hour (HTTP 200) | Shortest term a week: aime.info/en/gpucloud/ (the home page's link) answers 200 with nine plans (RTX 3090, RTX 4090, RTX 6000 Ada / L40S, A100 40/80GB, RTX PRO 5000/6000 Blackwell, H100 NVL, H200 NVL) priced weekly, monthly, half-yearly and yearly - "weekly 79,00 € corresponds to 0,47 € per hour". Every per-hour figure is the plan price divided by its hours, not an hourly rate a reader can rent; no plan is billed by the hour. | ||
| Akash Network | Not a GPU-hour (HTTP 200) | Prices are "derived from recent provider bids ... weighted by GPU count" - an average, which the marketplace rule excludes (a real ask, never an average). | — | |
| Aquanode | Not a GPU-hour (HTTP 200) | A static table of rates (RTX 3090 $0.500, A40 $0.490, L40 $0.772, RTX A4500 $0.250, RTX 4000 Ada $0.280) that the page itself calls the "cheapest current data-center offer ... aggregated across every data-center provider" - an aggregate of other operators' asks, not this seller's own price. | ||
| Aruba Cloud | Not a GPU-hour (HTTP 200) | The GPU on Demand pricing page answers 200 and prices two dedicated GPU servers by the month only: "Nvidia GeFORCE RTX 4080 Super ... € 529.90 + VAT / month" and "Nvidia L40S ... € 1979.90 + VAT / month", with "Free setup if you order for 6 months". No hourly figure and no stated hours-per-month divisor anywhere on the page, so there is no hourly rate to publish. | ||
| Cirrascale | Not a GPU-hour (HTTP 200) | Prices by the month, not by the hour, and says so: "All pricing shown for servers are per server per month". Every instance is offered on 1, 3, 6 and 12 month terms - 8x AMD MI300X at $22,499 monthly down to $17,999 annual, 4x MI250 at $4,679 monthly. The "$3.85/GPUhr Equivalent" beside each is Cirrascale's own division of a monthly price, not an hourly rate anyone can order. Readable and static; out of scope the way GreenNode and LeaderGPU are. | ||
| Cloudoye | Not a GPU-hour (HTTP 200) | GPU cloud servers priced by the month in rupees ("64 GB RAM - 16 vCore - 8 GPU Core From Rs 19,995.00 /mo"), with the card not named per plan. The one hourly figure is prose in an FAQ ("as low as Rs 40/hour for a devoted NVIDIA T4-based cloud GPU Instance"), a "from" sentence, not a rate card. | ||
| Cloudzy | Not a GPU-hour (HTTP 200) | GPU VPS priced by the month: RTX PRO 6000 $845/mo, RTX 5090 $649.35/mo. No hourly rate anywhere. | ||
| ComfyICU | Not a GPU-hour (HTTP 200) | Its pricing page prices GPU seconds in credits (L4 24GB $0.0009/s, L40S 48GB $0.0016/s, H100 80GB $0.0032/s), but "credits come with a monthly plan and don't roll over" - plans from $10 to $240 a month, each card available only from a given plan up. No GPU time is sold without a subscription, so no figure is the price of the hour alone. | ||
| compute.cheap | Not a GPU-hour (HTTP 200) | Prices set against a "Market" column - a reseller of other clouds' capacity, not an operator's own rate card. | — | |
| Database Mart | Not a GPU-hour (HTTP 200) | Prices GPU servers by the month only - "Advanced GPU VPS - RTX Pro 5000 $359.00/mo", the term selector starting at 3 months - and the page carries no hourly rate for any card. The site sells hourly nothing, so this is scope, not a layout we could read better. | ||
| DataPacket | Not a GPU-hour (HTTP 200) | Dedicated GPU servers by the month only: datapacket.com/dedicated-gpu-servers (linked from the home page) prints "Starting at $431 /mo" through "Starting at $17,202 /mo" per server (RTX PRO 6000, H200 NVL, HGX B300); no hourly rate. | ||
| Deepnote | Not a GPU-hour (HTTP 200) | A notebook product, not a GPU rate card: the pricing page offers "$50 worth of GPU every month" inside the Team plan, and the hardware docs (rendered) list "GPU (T4): 8 vCPU, 32 GB memory, 16 GB VRAM" and "GPU (L4)" machines with no price beside them and "contact us" for more. No hourly rate per card is published anywhere. | ||
| Digital Frontier | Not a GPU-hour (HTTP 200) | Whose GPUs the rates are is not stated: digitalfrontier.cloud/pricing prints two GPU rates ("NVIDIA T4 Tesla T4 $0.59 /hr", "NVIDIA A100 A100 40GB $4.05 /hr") under a compute table of three tiers - its own edge (Portugal, Finland, Bulgaria), "Hyperscaler GCP" ("runs on GCP data centers") and "Akash DePIN" reverse auctions - and the GPU rows name no tier; every other card is "priced live in the console via the Akash marketplace", i.e. other operators' offers. A reseller of GCP and Akash capacity is out of scope until a rate names its own hardware. | ||
| FDC Servers | Not a GPU-hour (HTTP 200) | Dedicated GPU servers by the month with a set-up fee: fdcservers.net/product/gpu-servers (linked from the home page) prints "NVIDIA L4/L40s/H100 ... From: $1,500 /month Set-up: $1,000" and "2/4/8 x NVIDIA L4/L40s/H100 ... From: $4,200 /month"; no hourly rate. | ||
| FireVM | Not a GPU-hour (HTTP 200) | GPU bare-metal servers rented by the month with days of setup, not by the hour. The page answers 200 and prices each server monthly with an hourly equivalent beside it - "Tesla T4 ... $349/month ($0.48/hour)", "H100 NVL ... $1999/month ($2.74/hour)", "H200 NVL ... $2499/month ($3.42/hour)" - each row "Within 5 business days" and "Order now". The hourly figure is the month over 730 hours; no server can be taken for an hour. Out of scope the way Leaseweb and LeaderGPU are. | ||
| FirstVDS | Not a GPU-hour (HTTP 200) | Not hourly: the GPU VDS page (linked from the apex) prices servers per day and only as "from" - "Nvidia L4, RTX 4090, RTX 5090 or L40S from 1 150 RUB/day" for a dedicated GPU, "from 299 RUB/day" for a vGPU slice of an L40S - with no rate per named card and no hourly figure anywhere on it. | ||
| FlexStack | Not a GPU-hour (HTTP 200) | A deployment platform on the customer's own AWS account, not a GPU cloud: its pricing page charges $0, $19 and $29 a month "+ AWS usage" and estimates CPU, memory and storage per hour; no GPU appears on it. | ||
| Fluence | Not a GPU-hour (HTTP 200) | Sells "GPU capacity ... across providers". The one figure on the page (H200 $2.80/hr) stands only inside a comparison against CoreWeave and AWS, not as a rate for a machine it sells. | ||
| GPUCloudHQ | Not a GPU-hour (HTTP 200) | Monthly rates only: H100 SXM $1,249/mo, H200 $1,690/mo, B300 $3,790/mo. No hourly figure and no stated term to divide a monthly price by. | ||
| GPULab | Not a GPU-hour (HTTP 200) | A JupyterLab notebook service rather than an hourly GPU cloud, and one card only: gpulab.io/pricing answers 200 with three plans for a dedicated NVIDIA K80 (12 GB) - $4.95 a day, $14.95 a week, $39.95 a month - and no hourly rate anywhere on the page. The plans are a flat fee for an immutable container, the enterprise option is "Contact us" for "advanced GPUs", and the footer is dated 2021. Priced by the day is not priced by the hour (the LeaderGPU and Database Mart reason), and the K80 is a card this site does not hold. | ||
| GPUonCLOUD | Not a GPU-hour (HTTP 200) | A broker of other operators' machines. Its pricing page, rendered, carries seven rates (L4 $0.03 an hour up to H100 80GB $2.50) under its own note that prices "may vary based on availability, region, and host provider", beside a link to "500+ GPUs in the marketplace with real-time pricing": indicative starting prices for machines other people own, not a rate card for a named machine - Runcrate's shape. The L4 at $0.03 is some twenty times under the median of every L4 this site reads, which is what an indicative floor looks like. | ||
| GPUOutlet | Not a GPU-hour (HTTP 200) | Rendered, the page is a "Reserved GPU Capacity Marketplace" - resale of reserved capacity, not an hourly rate card of its own. | — | |
| GreenNode | Not a GPU-hour (HTTP 200) | Its pricing page prices GPU instances per 30 days in Vietnamese dong, and prints hourly figures only for platform products ("Starting at 2,356 VND/hour"), never for a named card. Not a GPU-hour. The dong is also a rate our fx_rate column cannot carry to six decimals (about 0.000038 USD), so a dong source waits for that column to widen. Re-probed 2026-09-28. | ||
| HiAPI | Not a GPU-hour (HTTP 200) | Sells model API credits, not GPU time: the pricing page carries no GPU-hour figure. | ||
| HorizonIQ | Not a GPU-hour (HTTP 200) | Dedicated GPU servers by the month only: horizoniq.com/services/compute/gpu-dedicated-server/ (linked from the home page) prints three figures, "$1800* /month", "$1500* /month" and "$500* /month", and no hourly rate. | ||
| hosted.ai | Not a GPU-hour (HTTP 200) | A software vendor, not a GPU cloud: www.hosted.ai answers 200 with "Your GPU Commerce Platform - hosted.ai turns GPUs into AI cloud revenue for fleet owners, neoclouds, telcos and CSPs", selling GPUaaS software to operators ("Book a Demo"). No GPU is rented to a reader and no number stands beside a currency or an hour anywhere on the page. An aggregator lists it among GPU clouds. | ||
| Hosthink | Not a GPU-hour (HTTP 200) | Dedicated GPU servers by the month only: the GPU servers page lists each machine with one monthly price (1x RTX 4090 from $495/mo, 1x H100 80GB $1845/mo) and no hourly figure or stated hours-per-month divisor. | ||
| HOSTKEY | Not a GPU-hour (HTTP 200) | GPU servers priced by the month. Its GPU dedicated servers page answers 200 and lists 51 configurations, each with a monthly figure in euros and dollars ("GPU 1xRTX 4090 24GB EUR 339 /mo", "1x6000 PRO 96GB EUR 2190 /mo"); the page says servers are available on hourly and monthly plans, but the embedded product data marks every one "hourly_prices": false and no hourly rate is printed except a banner for an AMD R9700 server (a card with no id here). No hourly figure to read, and a month over an unstated divisor is not one - the case of LeaderGPU and Leaseweb. | ||
| Hostrunway | Not a GPU-hour (HTTP 200) | GPU servers priced by the month. Its GPU cloud server page answers 200 and prints every plan per month ("Quadro RTX A4000 ... $268 /mo", Quadro K620 $38 /mo); no hourly figure is printed, and a month over an unstated divisor is not one - the case of HOSTKEY and LeaderGPU. | ||
| Hydra Host | Not a GPU-hour (HTTP 200) | A procurement broker - "Wholesale pricing, enterprise procurement" - whose catalogue page carries no rate in the HTML and whose /pricing answers 404. | ||
| Impossible Cloud | Not a GPU-hour (HTTP 200) | Long-term cluster listings for future delivery (GB300 NVL72 "2027 Q1", 36-month terms, "Request this cluster") - capacity sold ahead, not an hourly rate. | ||
| ip-projects | Not a GPU-hour (HTTP 200) | Monthly only: ip-projects.de/en/dedicated-server/performance/gpu (the home page's GPU link) answers 200 with four dedicated GPU servers - "GPU Server S G2 €370.33 per month" to "GPU Server XL G2 €1522.44 per month" - and setup fees by minimum term (1, 6, 12, 24 months). No hourly figure; a server rented by the month is not an hourly GPU rate. | ||
| KAGOYA | Not a GPU-hour (HTTP 200) | Monthly only: kagoya.jp/cloudplatform/hpc/gpu/ (the home page's GPU link) answers 200 with four dedicated GPU servers priced in JPY as a setup fee and a monthly fee - "初期費用 902,000 円", "月額料金 858,000 円" down to "月額料金 97,900 円" - with "最低利用期間：2カ月" (minimum term two months). No hourly figure anywhere on the page; a server rented by the month is not an hourly GPU rate, and the month is never divided down by us. | ||
| KDDI GPU Cloud | Not a GPU-hour (HTTP 200) | Rents by the month, not the hour: its GPU cloud page prices the GB200 bare-metal server at "4,950,000円/月" (tax included) on the standard plan, and the 1- and 3-year plans read "お問い合わせ" (contact us). No hourly figure is published to read. | ||
| Lade | Not a GPU-hour (HTTP 200) | No GPU product. Its pricing page sells compute credits and CPU instances (from 0.0017 USD/hour, 1.25 USD/month upward) and names no GPU at all. | ||
| LeaderGPU | Not a GPU-hour (HTTP 200) | Rents by the period, not the hour: RTX 3090 "from EUR 125 / from EUR 249", 8x A40 EUR 238 / EUR 1,900, RTX 4090 EUR 329 a month. The bar is the term, not the currency - any currency has been read since 2026-09-28 - and no hourly figure is published to divide down from. | ||
| Leaseweb | Not a GPU-hour (HTTP 200) | GPU servers are dedicated servers billed by the month, not by the hour. The pricing page is an Angular shell, but the shop backend it names itself (shop-api.leaseweb.com, SAP Commerce OCC) answers our user agent with no key: 787 GPU offers, e.g. a Dell R750 (16SFF+8NVMe) GPU 2x Xeon 4314 with an L4 in Amsterdam at EUR 395.30. The product's own sapUnit is {"code":"MON","name":"Months"} and delivery is quoted at five days. Readable, but not an hourly rate - out of scope the way LeaderGPU is. | ||
| Mithril | Not a GPU-hour (HTTP 200) | The pricing page answers 200 and prints one column, "Spot (7-day avg)", per instance (H100 SXM5 x1 $1.51, x8 $21.00; H200 SXM5 x1 $3.93 and $2.15): "Mithril's omnicloud platform sets compute prices algorithmically based on supply and demand. Prices shown here are updated weekly and represent an average of the last 7 days." An average of past auction prices is not an ask a reader can pay, and there is no on-demand or reserved rate on the page. | ||
| NEOXIS | Not a GPU-hour (HTTP 200) | One dedicated server rented whole for a term, not an hourly rate card. Its pricing page answers 200 and prices a single 8x RTX PRO 6000 machine: "One fixed monthly price for the whole server", "about $9,930 a month for all 8 GPUs, excl. VAT", with a "Request a quote" button; the "from $1.70 per GPU-hour" beside it is that monthly figure divided down ("fixed for the term"), and the page sets it against per-hour offers that "change with demand". Nothing is billed by the hour. | ||
| NoctHost | Not a GPU-hour (HTTP 200) | A reseller of other clouds, not an operator: its home page says it deploys "on the infrastructure you already trust" - Vultr, DigitalOcean, Hetzner and Linode - from one prepaid balance, and "Exact pricing is shown at deploy time". Its GPU page sells A16 slices and A40 cards, so the rates are another provider's offer at a markup. | ||
| OneInfer | Not a GPU-hour (HTTP 200) | A broker of other clouds by its own words: its pricing page is a "Live catalog: 244+ models, 231+ GPU SKUs and 30+ providers", priced "per GPU-hour, with each provider's price side by side", and each GPU row reads "from $" - the lowest of other operators' offers, not a rate card of its own. | ||
| PoolCompute | Not a GPU-hour (HTTP 200) | A front over somebody else's marketplace, by its own words. Its home page carries one GPU-hour figure, "Deploy NVIDIA H100s at just $0.02/hr", under "Backed by The World's largest GPU Marketplace" and beside "connecting to largest pool" of underused GPUs - about 170x under our live H100 median of $3.47, an indicative floor for other machines, not a rate card. /gpu answers 200 with four card write-ups and no number beside a currency or an hour unit, before or after a render. Every "CHECK PRICES" link goes to marketplace.poolcompute.com, whose TLS handshake fails for requests and curl alike. | ||
| PowerGPU | Not a GPU-hour (HTTP 200) | Prices one card (H100 SXM $2.16 struck through to $1.512) under its own sentence "We buy capacity where the marketplaces do - then sell it": resale of marketplace capacity. | ||
| PowerPod | Not a GPU-hour (HTTP 200) | A front over a marketplace it does not run, by its own words: "the hosts belong to a marketplace and can go offline", and each "From, per hour" figure is the cheapest host of that card at the moment (H100 SXM 80 GB USD 2.78 /h, 2 hosts; RTX 4090 24 GB USD 0.65 /h). The table prints the cards and memory the way another marketplace lists them (L40S 45 GB, RTX 4070S Ti, Q RTX 8000), at about 1.5x the floors we read there (RTX 4090 0.65 vs 0.43, L40S 0.97 vs 0.60). A resold book, not a rate card of its own. | ||
| Prime Intellect | Not a GPU-hour (HTTP 200) | The home page lists the same card from several operators at once (H200 x1 at $1.99/HR, $1.80/HR and $1.23/HR) - a broker's offer list, not one seller's rate. No pricing page of its own (/pricing 404); the book is in the console at app.primeintellect.ai. | ||
| Qudata AI | Not a GPU-hour (HTTP 200) | A comparison site rather than an operator: its GPU pages count other providers' listings ("A100: 127 offers, 39 providers with A100"; "H100: 62 offers, 30 providers") and sell themselves as "Compare offers in one place, rather than checking dozens of provider websites". Not one currency figure appears on the catalogue page - every rate belongs to somebody else, and taking one would be copying a source we either read directly or have refused. The reason Runcrate and GPUonCLOUD are out. | ||
| Ravion | Not a GPU-hour (HTTP 200) | Not a GPU cloud: it is the successor to Flightcontrol, a deployment platform priced per module (compute stays in the customer's own cloud account). Its pricing page mentions GPUs zero times. | ||
| Runcrate | Not a GPU-hour | A broker across other clouds: each GPU is shown at the lowest of several offers from different providers and regions (its own API gives the minimum and maximum), so the figure is another operator's price, not a rate card of its own. | — | |
| RunDiffusion | Not a GPU-hour (HTTP 200) | Hosted image-generation apps sold by monthly subscription and tokens, not GPU time by the hour: rundiffusion.com/pricing answers 200 with plans "Runnit Hobby $8.79/mo", "Runnit Pro $23.99/mo", "Team $179.99/mo" and a reward balance, and no hourly figure or GPU model anywhere on the page. | ||
| Saturn Cloud | Not a GPU-hour (HTTP 200) | A platform over other operators' instances. Its Hosted plan page answers 200 and splits every rate into "Saturn Cloud" $0.00 and "Hosting" - "the charge for the underlying instances" - under the headings "Nebius Pricing" and "AWS Pricing": 1xH100 On-demand $2.95 (Nebius), T4-XLarge g4dn.xlarge $0.53, A10G-G5XLarge g5.xlarge $1.00 (AWS instance types named). Every figure is another cloud's price passed through, and Nebius and AWS are read directly. | ||
| Scan Cloud (Scan Computers) | Not a GPU-hour (HTTP 200) | Rents cloud workstations by the week, not the hour: the cloud solutions page (linked from scan.co.uk) lists products such as "1 Week - 32GB NVIDIA RTX 5090 ... £79.98" and ranges "From: £75.00", with no hourly figure anywhere on the page; the DGX Spark offer is per month with annual access. | ||
| Segmind | Not a GPU-hour (HTTP 200) | Its pricing page prints dedicated-endpoint GPU rates (H100 $9.00/hour, A100 $4.32/hour, L40 $2.66/hour, A40 $1.73/hour, A5000 $0.68/hour, A6000 $1.73/hour, or per second auto-scaling), but its plan table marks "Dedicated GPUs" unavailable on pay-as-you-go and included from the Pro plan ($39/mo) up; pay-as-you-go is priced per model call. No GPU hour is sold without a subscription. | ||
| Shadeform | Not a GPU-hour (HTTP 200) | A marketplace that states it has no rates of its own: "You pay the same as going direct to the clouds" in its marketplace. Nothing to read that is not already another provider's price. | ||
| Spheron | Not a GPU-hour (HTTP 200) | Its own pricing page calls the figures "live rates across 5+ providers" and "the cheapest live on-demand rate per GPU on the Spheron marketplace", with a "Supplier Matchmaking" service that sources from a partner network - a broker of other operators' offers, and a single cheapest ask rather than a floor over a window. | — | |
| Standard Compute | Not a GPU-hour (HTTP 200) | A flat LLM subscription (a $39 plan); no GPU rental and no GPU-hour rate on the site. | ||
| Synteq | Not a GPU-hour (HTTP 200) | Monthly only: synteq.com/gpu/cloud (the home page's link) answers 200 with a "Configure Your GPU Server" picker (Arc Pro A40, RTX PRO 4000, 5000, 6000, B300) whose one printed price is per month - "RTX PRO 6000 ... VRAM 96GB ... $ 885 /mo" - and no hourly figure anywhere in the page or its scripts. A dedicated server rented by the month is not an hourly GPU rate, and the month is never divided down by us. | ||
| Telnyx | Not a GPU-hour (HTTP 200) | A telecom API company whose GPUs serve its own inference product, not a GPU rental: telnyx.com answers 200, and its sitemap of 1,511 URLs carries 19 pricing pages (voice, messaging, numbers, storage, inference-api billed per token...) and none for a GPU instance. GPU-hour figures appear only in its blog articles about other providers' rates (an aggregator found it through "nvidia-b300-pricing"), which are not a price it charges. | ||
| Thalassa Cloud | Not a GPU-hour (HTTP 200) | A European Kubernetes and VM platform priced in euro per vCPU, per GB and per volume-hour; its pricing page names no GPU. | ||
| ThinkDiffusion | Not a GPU-hour (HTTP 200) | Hosted Stable Diffusion / ComfyUI machines at an hourly rate, but the machines are named by memory alone: thinkdiffusion.com/pricing answers 200 with "QUICK 16GB VRAM $0.99/hr, TURBO 24GB VRAM $1.75/hr, ULTRA 48GB VRAM $2.50/hr" (and 20%-discount rows) and names no GPU model anywhere on the page. A rate we cannot tie to one card is not a quote this site can list. | ||
| VALDI | Not a GPU-hour (HTTP 200) | Its GPU list (linked from valdi.ai) carries only capacity offers from third-party data centres with a minimum order in nodes and weeks - "8x H100 Tier 3 DC Des Moines, Iowa $2.18/GPU-hr ... Min. order: 8 nodes for 52 weeks", the shortest "1 node for 4 weeks" - no on-demand rate and no term we list; the on-demand page (/compute) prints no figure, even rendered. | ||
| Valebyte | Not a GPU-hour (HTTP 200) | Dedicated servers by the month only: valebyte.com/en/gpu-servers/ (linked from the home page) says "Tesla T4 starts from EUR 90/month, while RTX 4090 starts from EUR 1200/month" and "Dedicated Servers - from $155/mo"; no hourly rate. | ||
| VoltageGPU | Not a GPU-hour (HTTP 200) | Static rates (RTX 4080 $0.35/GPU/h, L40 $0.89/GPU/h, A100 $2.02) that the page says are "live from the Targon /inventory endpoint" across a "distributed provider network" - other operators' machines relayed, not its own rate card. | ||
| Wafer | Not a GPU-hour (HTTP 200) | An inference API, not GPU rental: no pricing page (/pricing 404), and MI350X is named only as hardware the service runs on. There is no machine to rent by the hour. | ||
| WebNX | Not a GPU-hour (HTTP 200) | Dedicated GPU servers by the month only: webnx.com/gpu-servers (linked from the home page) lists "NVIDIA RTX 5090 32GB ... $599 /mo", "4x NVIDIA V100 ... $899 /mo", "4x NVIDIA RTX PRO 6000 Blackwell 96GB ... $12,999 /mo" and no hourly rate anywhere on the page. | ||
| Wokey | Not a GPU-hour (HTTP 200) | Sells model API tokens, not GPU time: the home page lists per-1M-token input and output prices for hosted chat models beside their vendors' list prices, and no GPU, instance or hourly rate. | ||
| Yotta Data Services (Shakti Cloud) | Not a GPU-hour (HTTP 200) | Its static price list carries 209 figures in rupees and exactly one is per GPU per hour: "Shakti Cloud Azure ML studio - Azure ML Studio running on Shakti Cloud with H100 GPUs. 264 rupees / GPU / Hr", in a column headed "Price (Starting from)" under "Yotta - Microsoft AI Services". That is a bundled managed Azure service, not a GPU instance rate. Shakti's own GPU products are priced otherwise: "AI Lab-H100-10GB ... 10 GB slice of Nvidia H100 ... 30,000 rupees" is a fractional slice at a period rate. The rest is per core, per month or per million tokens. | ||
| ZenoCloud | Not a GPU-hour (HTTP 200) | An Indian managed-GPU host that prices by the month only. Its GPU hosting page answers 200 and prints every figure per card per month plus taxes - "NVIDIA L40S 48GB ... Rs 55,000 a month + taxes", RTX PRO 6000 Rs 1,10,000, H200 Rs 2,20,000; L4, A100, H100 and B200 "From" a monthly figure "On request" - beside managed-service fees per node per month. No hourly rate anywhere on the page. | ||
| Baseten · every GPU, on-demand one quote | Quote retired | Never serverless: the four figures come from "Dedicated Deployments" (GPU instances billed per minute) and the word "serverless" is absent from www.baseten.co/pricing/ - Baseten's serverless product is "Model APIs", priced per million tokens and not published by us. Republished with the market unset, so it follows price_type (on_demand), from the same figures. Price audit 2026-09-21. | — | |
| CoreWeave · NVIDIA A100 80GB, spot () one quote | Quote retired | CoreWeave spot is priced per region: the page prints one rate card for North America and one for Europe. This quote carried the cheaper of the two with an empty region, so the figure could not be found in either table by a reader following the link. Replaced by one quote per region from 2026-09-19; no figure was deleted. | — | |
| CoreWeave · NVIDIA B200, spot () one quote | Quote retired | CoreWeave spot is priced per region: the page prints one rate card for North America and one for Europe. This quote carried the cheaper of the two with an empty region, so the figure could not be found in either table by a reader following the link. Replaced by one quote per region from 2026-09-19; no figure was deleted. | — | |
| CoreWeave · NVIDIA B300, spot () one quote | Quote retired | CoreWeave spot is priced per region: the page prints one rate card for North America and one for Europe. This quote carried the cheaper of the two with an empty region, so the figure could not be found in either table by a reader following the link. Replaced by one quote per region from 2026-09-19; no figure was deleted. | — | |
| CoreWeave · NVIDIA H100 SXM, spot () one quote | Quote retired | CoreWeave spot is priced per region: the page prints one rate card for North America and one for Europe. This quote carried the cheaper of the two with an empty region, so the figure could not be found in either table by a reader following the link. Replaced by one quote per region from 2026-09-19; no figure was deleted. | — | |
| CoreWeave · NVIDIA H200, spot () one quote | Quote retired | CoreWeave spot is priced per region: the page prints one rate card for North America and one for Europe. This quote carried the cheaper of the two with an empty region, so the figure could not be found in either table by a reader following the link. Replaced by one quote per region from 2026-09-19; no figure was deleted. | — | |
| CoreWeave · NVIDIA L40, spot () one quote | Quote retired | CoreWeave spot is priced per region: the page prints one rate card for North America and one for Europe. This quote carried the cheaper of the two with an empty region, so the figure could not be found in either table by a reader following the link. Replaced by one quote per region from 2026-09-19; no figure was deleted. | — | |
| CoreWeave · NVIDIA L40S, spot () one quote | Quote retired | CoreWeave spot is priced per region: the page prints one rate card for North America and one for Europe. This quote carried the cheaper of the two with an empty region, so the figure could not be found in either table by a reader following the link. Replaced by one quote per region from 2026-09-19; no figure was deleted. | — | |
| Edgevana · every GPU, on-demand one quote | Quote retired | Fleet-index reads: the page's one lowestHourly per card was the cheapest available configuration, a whole multi-GPU machine when single GPUs were taken (H200 $55.70 against $5.89; audit different on the RTX 3090 Ti) and for the H200 a reserved server, published as one GPU on demand. No read states the GPU count, so none is converted. Read per card page with its GPU count from 11:40 UTC 2026-10-07 (method 1.32.25); owner's decision 2026-10-07. | — | |
| Edgevana · NVIDIA RTX PRO 4000 Blackwell, on-demand one quote | Quote retired | Mapped and published in error on 2026-10-04 against the owner's decision of 2026-09-27 (BACKLOG.md): the RTX PRO 4000 Blackwell is not listed until a source names the edition, and Edgevana's page and card page name none. One read at $0.31. | — | |
| GMI Cloud · NVIDIA H100 SXM, on-demand one quote | Quote retired | Never shown to be SXM: the GMI Cloud card reads "NVIDIA H100 GPU" with no form factor or memory, and h100-sxm rested on the research doc, not the page. Read as the form-factor-neutral h100 from 2026-09-19. | — | |
| ionstream · every GPU, on-demand one quote | Quote retired | ionstream no longer read: its CRO wrote (2026-09-28) that the "from" per-GPU-hour rates assume about 25% of the server paid up front and that the B200 is no longer available; the page states neither, so the rates cannot be published as on-demand. | — | |
| Meshive · NVIDIA A100 40GB, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA A100 40GB, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA A100 80GB, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA A100 80GB, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA A40, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA A40, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA B200, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA B200, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3060 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3060 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3060, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3060, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3070 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3070 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3070, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3070, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3080 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3080 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3080, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3080, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3090 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3090 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3090, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 3090, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4060 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4060 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4060, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4060, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070 SUPER, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070 SUPER, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070 Ti SUPER, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070 Ti SUPER, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4070, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4080 SUPER, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4080 SUPER, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4080, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 4080, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5060 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5060 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5060, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5060, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5070 Ti, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5070 Ti, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5070, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5070, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5080, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA GeForce RTX 5080, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA H100 PCIe, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA H100 PCIe, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA H100 SXM, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA H100 SXM, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA H200, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA H200, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA L4, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA L4, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA L40, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA L40, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA L40S, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA L40S, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 4090, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 4090, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 5090, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 5090, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 5880 Ada, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 5880 Ada, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 6000 Ada, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX 6000 Ada, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A2000, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A2000, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A4000, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A4000, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A5000, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A5000, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A6000, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX A6000, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX PRO 4500 Blackwell, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX PRO 4500 Blackwell, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX PRO 5000 Blackwell, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX PRO 5000 Blackwell, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX PRO 6000 Blackwell, on-demand () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Meshive · NVIDIA RTX PRO 6000 Blackwell, spot () one quote | Quote retired | The HTML price catalog included models outside Meshive's active fleet. Current rates come from the public API's gpu_offers array. | — | |
| Microsoft Azure · NVIDIA H200, spot one quote | Quote retired | Azure lists H200 spot at the on-demand price in every region (32 of 32, 2026-09-25): no spot market; parser emits spot only where it differs (commit 4f57fb0) | — | |
| Microsoft Azure · NVIDIA RTX PRO 6000 Blackwell, on-demand (uaenorth) one quote | Quote retired | Azure's price API listed this meter at two prices from 2026-10-01 ($0.01 and $7.87 per VM-hour); which one applies cannot be told from the data, so the meter is not read and the $0.01 read is not published. | — | |
| Microsoft Azure · NVIDIA RTX PRO 6000 Blackwell, spot (uaenorth) one quote | Quote retired | Azure's price API listed this Spot meter at two prices from 2026-10-01 ($0.002 and $1.574 per VM-hour); which one applies cannot be told from the data, so the meter is not read and the $0.002 read is not published. | — | |
| Nebius · every GPU, reserved one quote | Quote retired | Never a commitment: Nebius column 4 is "Preemptible, GPU-hour" (page and 2026-09-07 fixture). Read as committed until 2026-09-15, republished as spot from the same figures. Price audit 2026-09-15. | — | |
| packet.ai · every GPU, on-demand (Dynamic) one quote | Quote retired | packet.ai Dynamic is a shared card, read as market shared from 2026-09-27 (owner's decision); the on_demand quotes before that are superseded | — | |
| Paperspace · NVIDIA V100, on-demand (Paperspace) one quote | Quote retired | The Paperspace x8 node was "V100-32G", the 32 GB card, read as our v100 (the 16 GB card) from 2026-09-08; the parser reports it as unmapped from 2026-09-29, as immers and Cyfuture do. | — | |
| RunPod · every GPU, spot one quote | Quote retired | RunPod API returned spot equal to on-demand for every GPU and tier, 2026-09-07 to 2026-09-11; the parser emits spot only when it differs | — | |
| RunPod · NVIDIA RTX PRO 4000 Blackwell, on-demand (Secure Cloud) one quote | Quote retired | Mapped and published in error on 2026-10-04 against the owner's decision of 2026-09-27 (BACKLOG.md): the RTX PRO 4000 Blackwell is not listed until a source names the edition, and RunPod's API name does not. One read at $0.57 securePrice. | — | |
| Thunder Compute · NVIDIA H100 PCIe, on-demand one quote | Quote retired | Thunder Compute's redesigned pricing page (2026-09-29) names its H100 an "H100 SXM" at the same $3.20 it listed as "H100 PCIe" until 2026-09-28; the card is read as the page names it, as h100-sxm, from 2026-09-29. | — | |
| Vast.ai · NVIDIA A100 80GB, on-demand one quote | Quote retired | Never an A100 80GB: Vast gpu_name "A100 SXM4" covers the 40 GB and the 80 GB card and only gpu_ram tells them apart, so the floor was taken over both books and published the 40 GB machines (13 of 17 offers) as 80 GB - $0.5111 and $0.5891 per GPU against a real 80 GB floor of $1.13-$1.70. Parser now sends a gpu_ram window; the 40 GB book is a100-40. Price audit 2026-09-17. | — | |
| Vultr · every GPU, on-demand one quote | Quote retired | Vultr no longer read: its plans API (the only machine-readable price source) is disallowed for every agent by api.vultr.com/robots.txt, and the L40S plans it listed had zero locations - a rate nobody could order. 2026-09-18. | — | |
| Clore.ai · NVIDIA A100 40GB (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA GeForce GTX 1080 Ti one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA GeForce RTX 3060 Ti one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA GeForce RTX 4060 Ti (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA GeForce RTX 4070 SUPER (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA GeForce RTX 4080 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA RTX A2000 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA RTX PRO 6000 Blackwell one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA RTX PRO 6000 Blackwell (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Clore.ai · NVIDIA V100 32GB (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA A100 80GB (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA B200 one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA B200 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA B300 one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA B300 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA H100 PCIe one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA H100 SXM one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA H100 SXM (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA H200 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA H200 NVL one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA L40 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA L40S one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA RTX PRO 5000 Blackwell one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Lium · NVIDIA V100 32GB (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Octa.space · NVIDIA GeForce GTX 1660 SUPER one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Octa.space · NVIDIA GeForce RTX 5070 one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA B200 one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA B200 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA GeForce GTX 1660 Ti one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA GeForce RTX 2080 Ti (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA GeForce RTX 3070 Ti one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA GeForce RTX 3090 Ti (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA GeForce RTX 4080 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA H100 NVL one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA H100 NVL (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA H100 PCIe (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA H100 SXM (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA H200 one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA H200 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA L40 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA Quadro P4000 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA Quadro RTX 4000 one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 4000 Ada (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 4500 Ada Generation one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 4500 Ada Generation (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 5000 Ada Generation one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 5000 Ada Generation (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 5880 Ada one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 5880 Ada (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX 6000 Ada (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX A2000 one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA RTX A2000 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA Tesla P100 12GB (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 0 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA TITAN RTX (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 2 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA TITAN V one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA V100 one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — | |
| Vast.ai · NVIDIA V100 (Multi-GPU machines) one quote | Too few offers for a floor | Withheld: 1 verified offer(s), a floor needs 3. The floor is published again once enough offers are verified. | — |
Data updated: