> For the complete documentation index, see [llms.txt](https://opengpu-network.gitbook.io/opengpu-network/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://opengpu-network.gitbook.io/opengpu-network/for-clients/relay.md).

# Relay

Relay is an API-based AI service that uses OpenGPU Network as its infrastructure.

## What Is Relay?

Relay provides a developer-friendly API gateway to GPU compute—no blockchain knowledge required.

* **Familiar APIs**: Compatible with standard model interfaces
* **No wallet needed**: Standard API key authentication
* **Multiple AI capabilities**: Chat, TTS, ASR, image generation
* **Dual-mode**: Can use OpenGPU Network (decentralized) or direct relaying (central routing)

## When to Use Relay vs OpenGPU

| Use Relay                          | Use OpenGPU                |
| ---------------------------------- | -------------------------- |
| Want simple API access             | Want full protocol control |
| Don't need blockchain transparency | Need on-chain verification |
| Prefer API keys over wallets       | Building custom Sources    |
| Real-time, low-latency needed      | Building on the protocol   |

## Quick Example

```bash
curl -X POST https://relay.opengpu.network/api/ollama/chat \
  -H "Content-Type: application/json" \
  -H "X-API-Key: YOUR_API_KEY" \
  -d '{
    "model": "llama3.2:3b",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

## Getting Started with Relay

1. Visit [relaygpu.com](https://relaygpu.com)
2. Try as guest or sign up for free credits
3. Start making API calls

Relay handles all blockchain interactions to abstract away web3.

→ **Full Relay documentation**: [docs.relaygpu.com](https://docs.relaygpu.com)