Quickstart

Make your first LibertAI call in under five minutes. If you already use OpenAI or Anthropic SDKs, you just change the base URL.

1. Get an API key

Sign in to the Developer console, create a key, and top it up. Or skip the key entirely and pay per request in USDC on Base — see x402 payments.

2. Make a call

curl -X POST https://api.libertai.io/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $LIBERTAI_API_KEY" \
  -d '{
    "model": "gemma-4-31b-it",
    "messages": [
      { "role": "user", "content": "In one sentence: what is decentralized inference?" }
    ]
  }'

That's it. You're talking to an open-source LLM on a decentralized compute network.

Already using OpenAI?

Two lines change. Keep the rest of your code.

- base_url="https://api.openai.com/v1"
+ base_url="https://api.libertai.io/v1"
- model="gpt-4o-mini"
+ model="gemma-4-31b-it"

The /v1/chat/completions, /v1/completions, /v1/responses, and /v1/models endpoints are all OpenAI-compatible. /v1/messages is Anthropic-compatible.

Already using Anthropic?

- base_url="https://api.anthropic.com"
+ base_url="https://api.libertai.io"
- model="claude-haiku-4-5"
+ model="gemma-4-31b-it"

The /v1/messages endpoint accepts the Anthropic Messages format directly.

Next steps

• Use LibertAI in your tools — Cline, Roo Code, Continue, OpenCode, Aider, Claude Code, LangChain
• Text API — list of models, pricing, context windows, capabilities
• Text usage — streaming, function calling, vision, direct CRN calls
• Image generation — Stable Diffusion and OpenAI-compatible endpoints
• Web search — multi-engine search and URL fetching for RAG
• Agents — build autonomous agents that pay for their own inference
• x402 payments — call the API with crypto instead of an API key
• Architecture — how requests are routed across the network