> ## Documentation Index
> Fetch the complete documentation index at: https://docs.compute-desk.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Compute Clear API

> The most comprehensive API for GPU compute across cloud providers

The Compute Clear API gives you access to the broadest selection of GPU compute in the market — 20+ cloud providers through a single API. Discover real-time pricing, provision instances, and manage resources across Nebius, Lambda Labs, Verda, Hyperstack, Scaleway, Crusoe, MassedCompute, DigitalOcean, and many more.

Every offer is validated against your requirements through our proprietary verification layer — you define what you need in a [YAML spec](/guides/yaml-spec), and we guarantee that what you procure matches. No surprises.

## What you can do

<CardGroup cols={2}>
  <Card title="Discover GPUs" icon="magnifying-glass" href="/quickstart">
    Query real-time pricing and availability across all vendors
  </Card>

  <Card title="Procure Instances" icon="server" href="/guides/offer-lifecycle">
    Provision GPU instances with SSH access and cloud-init
  </Card>

  <Card title="Manage Resources" icon="sliders" href="/api-reference/endpoint/list-resources">
    List, monitor, and destroy your provisioned resources
  </Card>

  <Card title="Compare Pricing" icon="chart-line" href="/api-reference/endpoint/get-offers">
    Find the cheapest GPU for your workload across all providers
  </Card>
</CardGroup>

## Supported vendors

Nebius, Lambda Labs, Verda, Hyperstack, Scaleway, MassedCompute, Crusoe, DigitalOcean, Paperspace, Vultr, Denvr, Horizon, IMWT, ExcessSupply, Vast.ai, Sesterce — and more being added regularly.

Use [`GET /available_vendors`](/api-reference/endpoint/available-vendors) for the current list, and [`GET /available_gpus`](/api-reference/endpoint/available-gpus) to see which GPU types are available.

## Quick example

```bash theme={null}
curl -X POST https://supply-api.compute-desk.com/get_offers \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/yaml" \
  -d '
compute:
  gpu_type: H100
  gpu_count: 1
  max_price_per_gpu_hour: 5
'
```