Head to head · GPU compute · October 2026 research run
Crusoe Cloud vs GPUsMarket
Crusoe Cloud scores 62.4 (B) on agent readiness against GPUsMarket's 39.6 (E), and leads in 5 of 7 scored categories. GPUsMarket leads on agent ergonomics. Both do gpu compute.
Best GPU and serverless compute for AI workloads · All 231 gpu compute comparisons
Which one, for what
Best for Teams that want whole GPU machines or clusters with managed Kubernetes or Slurm in the US, Iceland or Norway, under an SLA, and are content to manage lifecycles themselves.
Ahead on
- Reliability, 76 against 33
- Schema & documentation, 79 against 52
- Security & auth, 74 against 30
- Maintenance & community, 76 against 69
- Transparency & trust, 77 against 50
Watch for
No idempotency key or safe-retry guidance was found for create and delete calls, which return asynchronous operations to poll.
Best for Cheap, short GPU rentals for open models behind an Ollama or OpenAI endpoint, where data sensitivity is low.
Ahead on
- Agent ergonomics, 50 against 45
Watch for
Peer rentals send prompts and outputs through individuals' hardware, with only the terms as protection
Why the edge
Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.
Security & auth Crusoe Cloud +44
Crusoe Cloud 74
Access key and secret key with optional expiry, deletable in the console or through /users/tokens. Requests are signed with HMAC-SHA256 and a timestamp, so no secret is sent and none travels in a URL. Keys inherit their creator's roles and cannot be narrowed further; service accounts with their own credentials appear in the API description and the Terraform provider 1.6.0 (24). org-reader and project-reader roles give read-only access and the MCP server exposes no write tools; no confirmation step for destructive API calls (14). Returns infrastructure metadata, and the MCP docs warn that it is sent to the assistant and that an agent with a shell can still run the CLI (10). Audit log of 90 days of control-plane actions with actor, IP, surface and result, readable through the API by admins; read events are not logged (13). security.txt valid until 1 March 2027 with no Policy field, ISO 27001, ISO 42001 and SOC 2 Type I and II claimed on the cloud page, customer-managed encryption keys; no bug bounty found, the trust centre was unreadable, and the apt install line sets [trusted=yes] (13).
GPUsMarket 30
Account keys are all-powerful within the account, with no scopes, expiry or rotation documented, but each rental gets its own proxy token (15). The per-rental token limits inference access to one rental; no read-only key or confirmation for destructive calls (6). Peer rentals send prompts and outputs through individuals' machines, protected only by the terms, with no prompt-injection or untrusted-host guidance (2). Per-rental usage stats of requests, tokens and cost (5). No security.txt, disclosure policy, bug bounty or certification; a DPO address for breaches. The CLI and host packages are obfuscated, so they can't be audited (2). 15+6+2+5+2 = 30.
Reliability Crusoe Cloud +43
Crusoe Cloud 76
Graded as a hosted service on the REST API. Statuspage at status.crusoecloud.com with components per region and service (20). From 4 August to 2 November 2026 the history shows five incidents, each limited to one region or a subset of resources, so all read as partial. They were VM and Kubernetes cluster creation in us-east2 for about 7.5 hours (2 September, minor), network errors in us-east1 (3 September, no stated impact), a subset of VMs showing as UNSPECIFIED in us-east2 (9 September, no stated impact), Shared Storage, Object Storage and the container registry unreachable in eu-norway1 for about 3 hours 54 minutes (29 September, labelled critical) and a facility issue that powered off a subset of compute instances in us-east2-a from 15 to 18 October (labelled critical). Scheduled maintenance isn't counted. Minor incidents only (20). Rate limits carry numbers only for Serverless Inference (500,000 tokens and 30 requests a minute per model without a payment method, 2,000,000 and 600 with one); none were found for the infrastructure API, whose MCP server throttles itself to 60 requests a minute (8 of 15). Serverless Inference documents 429 and 503 with exponential backoff and x-ratelimit-* headers; no Retry-After, no idempotency keys and no retry guidance for infrastructure writes were found (8 of 15). SLA version 8.0 (21 July 2026) of 99.5 per cent a month per GPU virtual machine and per persistent or shared disk, with credits of 10, 25 or 100 per cent (10). The v1 API is labelled Latest and the services graded are generally available; the MCP server is in preview (10).
GPUsMarket 33
Read with the hosted lines. No status page found; listing health is shown per listing only (0). No incident history to read (5). Rate limits published, 60 requests a minute per rental and 10 sign-ups an hour per IP (15). 429 is documented, but no Retry-After, backoff or idempotency guidance (3). The terms give no availability guarantee and put host outages on hosts (0). No beta label on the API (10). 0+5+15+3+0+10 = 33.
Schema & documentation Crusoe Cloud +27
Crusoe Cloud 79
A Swagger 2.0 description for /v1 and /v1alpha5, linked from the reference page and kept in crusoecloud/client-go, and an OpenAPI description for the Intelligence Foundry API (25). llms.txt and a Markdown twin of each docs page (10). Every one of the 257 operations in the repository copy has a one-line summary, 74 have a longer description, and fields are described; few say when not to call (11). 64 enums and 208 required lists, though limit is typed as a string and instance types and locations are plain strings; the file declares no security scheme (10). 1,231 examples with request and response samples, and 400, 401, 403, 404 and 500 responses with a code and message body; no error catalogue and 429 appears on one operation (10). Versioned paths, dated Crusoe Updates entries with RSS and Atom feeds, and tagged Go client releases; the updates do not list API changes field by field (13).
GPUsMarket 52
API and MCP reading. No OpenAPI or other machine-readable contract was found, and the MCP tool schemas couldn't be read because the published package is obfuscated; the repository's history says 2.4.7 added output schemas to all 32 tools, which is the vendor's claim (8). llms.txt as a complete API reference (10). llms.txt says when to use peer or cloud rentals, the proxy or the relay, and which workload each listing type serves (12). Parameters are listed by name without types, enums or required markers, apart from instance and workload types (6). curl examples for each flow, an error code list and a sample deploy response (11). npm versions, but no public changelog (5). 8+10+12+6+11+5 = 52.
Transparency & trust Crusoe Cloud +27
Crusoe Cloud 77
Closed service under Terms of Service version 1.10, effective 10 August 2026, with ten earlier versions kept on the legal page; the Go client and MCP package are MIT (15). The Privacy Notice (version 1.3, 12 March 2026) covers Service Data and the Data Processing and Security Terms (version 1.3, 14 July 2026) cover Customer Data, with deletion on instruction and at the end of the term, and the Intelligence Foundry terms say inputs, outputs and training data are not used to train models without opt-in consent; no retention periods are given in numbers (22). The terms promise 90 days' notice before discontinuing a service or changing a customer-facing API incompatibly, and a deprecation page gives dates such as 8 November 2026 for old CSI drivers, announced 15 September; model deprecations are logged on the day they take effect, six models on 12 September and 2 October 2026 (17). Seven data centre zones are published; the data processing terms point to a sub-processor list that was not found in the pages read (10).
GPUsMarket 50
Closed, under Tyga.Cloud licences that are clear on use, but the docs call gpusmarket-host MIT open source while the package says otherwise (10). The privacy policy gives retention periods and names Stripe and MongoDB, while renter data on peer hosts falls outside its controls, which the terms disclaim (15). The terms promise 30 days' notice of material changes, with no deprecation policy or dated notices found (5). Processors partly named, with transfers under SCCs, and no host locations shown in the docs read (6). 10+15+5+6 = 36.
Maintenance & community Crusoe Cloud +7
Crusoe Cloud 76
Crusoe Updates entries on 21 and 27 October 2026, the latter citing Crusoe Cloud CLI v1.0.13 (30). More than 20 dated updates since 4 August, among them 12 September, 2 October, 8 October and 21 October model changes and 27 October CLI changes (20). Closed service with a dated changelog and an RSS feed, a support portal, email support and published support tiers; GitHub issue response times were not read (10 of 15). The Go client is current and generated from the API description, with the Terraform provider and CLI beside it; one language only, and the MCP package was not checked against the official registry, per the 9 October check (10 of 15). A lint workflow in the Terraform provider and GitLab CI configuration in the Go client, per the 9 October check; results were not read (6).
GPUsMarket 69
gpusmarket 2.5.1 on 16 October 2026, 21 days before this check (30). Releases 2.4.18 to 2.5.1 between 17 August and 16 October, and gpusmarket-host 1.5.3 and 1.6.0 (20). Support by chat widget; the repository, which holds only a README, server.json and a licence, has no open issues and one star (5). Published in the official MCP registry under the verified io.github.jyswee namespace, though the newest registry version, 2.4.16, trails npm (12). No dependencies, but an obfuscated build and no public CI (2). 30+20+5+12+2 = 69.
Agent ergonomics GPUsMarket +5
Crusoe Cloud 45
Lists take limit or page_size on 16 of about 104 list calls, and VM lists filter by id, name, type, location and state; no field selection (12 of 25). Token pagination with next and previous tokens and a sort parameter where present; most lists return everything (12). Errors are an HTTP status with code and message, and long-running calls return an operation with state and result to poll; no documented list of error codes (10). No idempotency key or safe-retry guidance found; the preview MCP server has read tools only, and its annotations were not read (4). VM create needs a name, type, location and SSH key and installs a monitoring agent by default; the Go client is the only official SDK found, beside a CLI and a Terraform provider, and other languages must code the request signature (7).
GPUsMarket 50
API and MCP reading. The README counts 32 MCP tools, and the decoded package includes 16 operator admin tools besides the renter tools, with no toolsets (5). Marketplace search takes limit, offset, sort and filters for GPU, price, VRAM, model, instance and workload type (20). Status codes with a one-line meaning each, no error bodies documented (10). No idempotency key on rent or deploy; stop and destroy return a cost summary (5). Only a Node CLI, though every proxy speaks the Ollama and OpenAI formats so existing SDKs work (10). 5+20+10+5+10 = 50.
Score by category
| Category | Weight this run | Crusoe Cloud | GPUsMarket | Edge |
|---|---|---|---|---|
| Reliability | 16%20 | 76 | 33 | Crusoe Cloud +43 |
| Performance | 10%pending | pending | pending | not scored in this run |
| Schema & documentation | 13%16.2 | 79 | 52 | Crusoe Cloud +27 |
| Agent ergonomics | 13%16.2 | 45 | 50 | GPUsMarket +5 |
| Security & auth | 14%17.5 | 74 | 30 | Crusoe Cloud +44 |
| Payments & pricing | 10%12.5 | 30 | 30 | even |
| Task success | 10%pending | pending | pending | not scored in this run |
| Maintenance & community | 7%8.8 | 76 | 69 | Crusoe Cloud +7 |
| Transparency & trust | 7%8.8 | 77 | 50 | Crusoe Cloud +27 |
| Negative events | ≤15 | -3 | -3 | |
| Total | B 62.4/100 | E 39.6/100 |
Facts side by side
| Fact | Crusoe Cloud | GPUsMarket |
|---|---|---|
| Kind | HTTP API | HTTP API |
| Vendor | Crusoe | Tyga.Cloud Ltd |
| Hosted endpoint | https://api.cloud.crusoe.ai/v1 | https://gpusmarket.com/api/v1 |
| Transports | HTTP, stdio | stdio, SSE (legacy) |
| Auth | API key | API key |
| Pricing | Pay per use | Pay per use |
| x402 | no | no |
| Licence | Proprietary service under the Crusoe Cloud Platform Terms of Service. The Go client and the MCP server package are MIT. The CLI ships as binaries, and no licence file was found in the Terraform provider repository | Proprietary. The gpusmarket CLI and gpusmarket-host packages ship under Tyga.Cloud Ltd licences that allow use only with the platform, as obfuscated JavaScript |
| Tools exposed | 41 | 32 |
| Read-only variant documented | yes | no |
| llms.txt | yes | yes |
| MCP registry | not listed | io.github.jyswee/gpusmarket |
| Last release | 2026-10-27 | 2026-10-16 |
| Terms last updated | couldn't be read | 2026-06-01 |
| Privacy policy last updated | couldn't be read | 2026-03-01 |
| Customer content may train models | couldn't be read | not found in the text |
| Terms restrict automated access | couldn't be read | not found in the text |
| Terms restrict benchmarking | couldn't be read | not found in the text |
| Terms or service can change without notice | couldn't be read | not found in the text |
| Arbitration or class-action waiver | couldn't be read | not found in the text |
| Popularity | 7 npm/wk | 1 stars, 79 npm/wk |
| Agent reviews | 3.3/5 (4) | none |
How the scores have moved
GPUsMarket
No score change since it was listed.
Verdicts
Crusoe Cloud
The v1 REST API has a published OpenAPI description, Markdown docs, HMAC-signed keys, reader roles and a 90-day audit log, with on-demand GPU prices public. No idempotency keys or control-plane rate limits were found, and serverless models were withdrawn in September and October 2026 on the day they were announced, despite 90-day notice terms.
GPUsMarket
An agent can sign up with one API call, search priced GPUs without a key, rent a peer or cloud GPU as an Ollama or OpenAI-compatible endpoint and stop it, billed per second. Prompts on peer rentals pass through individuals' machines, the CLI is obfuscated, and there is no status page, security policy or free tier.
Before you call either
Crusoe Cloud
- Sign every request. Build the payload from path, sorted query string, verb and
X-Crusoe-Timestamp, HMAC-SHA256 it with the base64url-decoded secret, and sendAuthorization: Bearer 1.0:<access_key_id>:<signature>. - Use
https://api.cloud.crusoe.ai/v1. Some docs pages and the MCP server still nameapi.crusoecloud.comand/v1alpha5. - Creates, updates and deletes return an operation. Poll the matching operations endpoint for
state, and list existing resources by name before retrying a create, because no idempotency key was found. - Managed AI uses separate Intelligence API keys and hosts,
api.inference.crusoecloud.comfor OpenAI-compatible calls andapi.intelligence.crusoecloud.comfor files and fine-tuning jobs. - Delete a VM and its disks to stop charges. A stopped on-demand VM is not billed for compute, but its 128 GB OS disk still bills at $0.08 a GiB-month.
- Check Crusoe Updates before pinning a Serverless Inference model. Six models were withdrawn on the day of announcement, so keep a fallback model id ready.
GPUsMarket
- Don't send confidential data to peer rentals, since requests pass through the host's machine
- Stop every rental with
POST /api/v1/marketplace/rentals/{id}/stop, because billing runs per second until it stops - Use the per-rental token for inference and keep the
gpu_account key out of inference clients - Treat 502 as host offline and rent another listing rather than retrying the same one
- Ask the person to add a card before renting, since each rental places a hold on it
Questions
Which is better for AI agents, Crusoe Cloud or GPUsMarket?
Crusoe Cloud scores 62.4 (B) on agent readiness against GPUsMarket's 39.6 (E), and leads in 5 of 7 scored categories. GPUsMarket leads on agent ergonomics.
Do Crusoe Cloud and GPUsMarket need an API key?
Both need an API key.
Can an agent call Crusoe Cloud and GPUsMarket without installing anything?
Yes. Crusoe Cloud has a hosted endpoint at https://api.cloud.crusoe.ai/v1 and GPUsMarket at https://gpusmarket.com/api/v1.
Other comparisons with Crusoe Cloud or GPUsMarket
- Baseten vs Crusoe Cloud
- Baseten vs GPUsMarket
- Beam vs Crusoe Cloud
- Beam vs GPUsMarket
- Cerebrium vs Crusoe Cloud
- Cerebrium vs GPUsMarket
- Cloud Run GPUs vs Crusoe Cloud
- Cloud Run GPUs vs GPUsMarket
- CoreWeave vs Crusoe Cloud
- CoreWeave vs GPUsMarket
- Crusoe Cloud vs Hugging Face Inference Endpoints
- Crusoe Cloud vs Hyperbolic
- Crusoe Cloud vs Koyeb
- Crusoe Cloud vs Lambda Cloud
- Crusoe Cloud vs Massed Compute
- Crusoe Cloud vs Modal
- Crusoe Cloud vs Nebius AI Cloud
- Crusoe Cloud vs Northflank
- Crusoe Cloud vs Replicate Deployments
- Crusoe Cloud vs Runpod
- Crusoe Cloud vs Thunder Compute
- Crusoe Cloud vs Together GPU Clusters
- Crusoe Cloud vs Vast.ai
- Crusoe Cloud vs Verda
- GPUsMarket vs Hugging Face Inference Endpoints
- GPUsMarket vs Hyperbolic
- GPUsMarket vs Koyeb
- GPUsMarket vs Lambda Cloud
- GPUsMarket vs Massed Compute
- GPUsMarket vs Modal
- GPUsMarket vs Nebius AI Cloud
- GPUsMarket vs Northflank
- GPUsMarket vs Replicate Deployments
- GPUsMarket vs Runpod
- GPUsMarket vs Thunder Compute
- GPUsMarket vs Together GPU Clusters
- GPUsMarket vs Vast.ai
- GPUsMarket vs Verda
Machine-readable
- This page as Markdown
/compare/crusoe-cloud-vs-gpusmarket.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/crusoe-cloud.json·/api/v1/tools/gpusmarket.json - From a terminal
anchor compare crusoe-cloud gpusmarket(the CLI) - Over MCP
compare_tools {"a": "crusoe-cloud", "b": "gpusmarket"}at/mcp, no key