# Hyperbolic (slim)

> Hyperbolic is a GPU cloud from Hyperbolic Labs that rents H100, H200 and B200 virtual machines and bare-metal nodes by the hour, with reserved terms and private clusters. Agents use its REST API with an API key.

- Full: https://www.anchorterminal.com/tools/hyperbolic.md (~8,400 tokens) · this version ~1,780 tokens · JSON https://www.anchorterminal.com/tools/hyperbolic.json · canonical https://www.anchorterminal.com/tools/hyperbolic
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-09

**D · 48/100 · rank #730 of 842 · #15 in GPU & serverless compute · not agent-ready · confidence medium**


Assessment: The API serves its own OpenAPI 3.1 document without a key, lists GPU prices on an open endpoint and dates the removal of legacy paths at 1 March 2027. API keys carry no scopes and can delete the account, rental creation has no idempotency key, and renting needs a browser signup and a $5 deposit.

## Facts

- Kind: HTTP API · vendor: Hyperbolic Labs, Inc. · category: GPU & serverless compute · legal entity: Hyperbolic Labs, Inc. · provenance 76/100
- Endpoint: `https://api.hyperbolic.ai` (HTTP)
- Auth: API key · pricing: Pay per use · x402: no · licence: Proprietary service under Hyperbolic's Terms of Service. The unmaintained hyperbolic-mcp repository on GitHub is MIT
- Probe metrics: not measured yet (probes haven't run)
- Surface graded: The REST API at https://api.hyperbolic.ai (OpenAPI 3.1, version 2.0.0, 65 operations, 31 of them deprecated aliases or per-type endpoints). The same API answers on the legacy host api.hyperbolic.xyz
- Operations: Create, get, list, update, terminate and reboot rentals. Convert to and extend reserved terms with quotes. Create, attach, resize and terminate storage volumes. List and manage SSH keys. Read the balance, the user and the API key list
- MCP server: None in the docs. HyperbolicLabs/hyperbolic-mcp (MIT, stdio, 9 tools) last changed on 21 May 2025 and targets the legacy v1 marketplace paths
- Free tier: An account is free, but renting needs a one-time $5 deposit. Free-tier API limit 60 requests a minute
- GPUs: `h100`, `h200`, `b200` and `rtx-6000-pro` in the API. The docs also list B300, which the API's `gpuType` enum does not include. Virtual machines or bare-metal nodes of 8 GPUs, with Ethernet or InfiniBand
- Scale to zero: None. Rentals are whole machines that bill until terminated. No serverless endpoint or job API in the reviewed documentation
- Cold start: The docs say an instance is ready in under 5 minutes. Instances not ready within about 3 hours are terminated and not charged
- Billing basis: On-demand hourly from ready, at a rate fixed when the rental is created. Reserved paid up front for 7, 14 or 30 days. Storage volumes hourly per GB of capacity
- Rate limits: 60 requests a minute on the free tier and 600 on the pro tier, per the plan cards in the docs. No 429 response or retry header is documented
- Errors: `code`, `message` and an `issues` array. Top-level codes BAD_REQUEST, UNAUTHORIZED, FORBIDDEN, NOT_FOUND and INTERNAL_SERVER_ERROR, with named cases such as RENTAL_NOT_FOUND and INVALID_INPUT_FOR_RENTAL_TYPE in operation descriptions
- Regions: 30 region codes in the create call's enum across the United States, Canada, Europe, the United Kingdom, the Middle East and Japan. Capacity comes from what the docs call an aggregated supplier network
- SDKs: None found for the v2 GPU API. The `@hyperbolic/ai-sdk-provider` npm package targets model inference
- Roles: Organisation admin and member. Both can launch and terminate instances and create keys. Only admins handle deposits, billing history and membership
- Status: status.hyperbolic.ai on UptimeRobot, with 90-day uptime per monitor. The page loads its data by script and we couldn't read it
- Support: In-app chat and support@hyperbolic.ai. Engineers work Monday to Friday, 06:00 to 17:00 Pacific Time
- Prices: H100 SXM5 80 GB virtual machine, on-demand $3.19 per GPU-hour

- Scores: Reliability 52, Performance pending, Schema & documentation 77, Agent ergonomics 55, Security & auth 47, Payments & pricing 15, Task success pending, Maintenance & community 42, Transparency & trust 61 · negative events -3 · total over the 7 assessed categories
- Why: Reliability, Graded on the hosted lines for the REST API. · Schema & documentation, Public OpenAPI 3.1 document with 65 operations and 53 schemas, served by the API without a key and matching the docs copy (25). · Agent ergonomics, No field selection. · Security & auth, Plain API keys sent as a Bearer header, shown once, revocable, any number per account, and attributed to their creator inside an Organisatio… · Payments & pricing, No x402, MPP or L402 on the GPU API. · Maintenance & community, The newest changelog entry is for the week of 5 October 2026 (30). · Transparency & trust, Closed service under Terms of Service dated 24 March 2025 and California law.
- Sources: 22, open questions: 10, both in the full twin
- Capabilities: compute.gpu
- JSON: https://www.anchorterminal.com/api/v1/tools/hyperbolic.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/hyperbolic.svg` or a link to https://www.anchorterminal.com/tools/hyperbolic from a page on hyperbolic.ai or one of its subdomains, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.

## Before you call it

1. Read `GET /v2/on-demand/rental-options` first. It needs no key and lists what can be rented now, with `costPerHourCents` per GPU configuration.
2. Send `rentalType` and `gpuCount` to `POST /v2/on-demand/rentals`. Region defaults to `us-central-1` and GPU type to `h100`, so set both from the options list.
3. List active rentals before retrying a failed create call. There is no idempotency key and every ready rental bills hourly.
4. Save an SSH public key with `POST /v2/ssh-keys` before renting. Without `sshPublicKeyIds` the newest saved key is attached.
5. Do not create reserved rentals unless asked. They are paid in full up front and cannot be terminated early.

## Connect

```bash
curl -X POST https://api.hyperbolic.ai/v2/ssh-keys -H "Authorization: Bearer $HYPERBOLIC_API_KEY" -H "Content-Type: application/json" -d "{\"publicKey\": \"$(cat ~/.ssh/id_ed25519.pub)\", \"name\": \"Work laptop\"}"
```
Full config and headless snippets are in the full page. Through letme (picks today, calling later): https://letme.dev/hyperbolic

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| Nebius AI Cloud | B | 67.2 | compute.gpu | https://www.anchorterminal.com/tools/nebius-ai-cloud.min.md |
| Baseten | B | 66.5 | compute.gpu | https://www.anchorterminal.com/tools/baseten.min.md |
| Hugging Face Inference Endpoints | B | 64.5 | compute.gpu | https://www.anchorterminal.com/tools/hugging-face-inference-endpoints.min.md |
| Modal | B | 63.6 | compute.gpu | https://www.anchorterminal.com/tools/modal.min.md |
| Replicate Deployments | B | 63.6 | compute.gpu | https://www.anchorterminal.com/tools/replicate-deploy.min.md |


## Panel reviews (0, desk reviews from public material, no calls made)