API overview

Base URLCopy Anchor Link

https://binference.io/api/v1

For the Anthropic SDKs and Claude Code, use https://binference.io/api: they add /v1 themselves.

EndpointsCopy Anchor Link

Any other path under /api/v1, like embeddings, answers 404 not_found in JSON.

DEX market dataCopy Anchor Link

The DEX's markets, order books, trades and candles, under the same base URL. They're public: no key needed.

• Numbers as decimal strings. Prices are USDT per $1 of credit, sizes are dollars of credit and USDT amounts are USDT. Times are Unix milliseconds.
• Cached for 1 second. Ask at most once a second.
• 1,200 weight a minute per IP address. Each call counts 1 to 5. Over it, the answer is 429 rate_limited with the seconds to wait in Retry-After.
• Only the query parameters each endpoint lists, each once. Any other is refused with 400 unknown_parameter.
• Errors are JSON: { "error": { "code": "unknown_market", "message": "..." } }. 404 not_found means the DEX is not open yet.
• Wallets and accounts never appear.

MCP serverCopy Anchor Link

Agents can also read their credit, usage and limits, and manage their own keys, over MCP at https://binference.io/api/mcp, with the same key. See MCP server to connect, and MCP tools for every tool.

AuthenticationCopy Anchor Link

Send your binf_ key as a bearer token or as x-api-key:

Authorization: Bearer binf_...

A missing, wrong or revoked key answers 401 invalid_api_key. See API keys.

RequestsCopy Anchor Link

• JSON bodies, up to 4 MB. Send images and files as URLs, not inline data: upload a large image and send its link.
• Bodies pass through as sent. Every field of each format works, beyond the ones these pages list.
• Stateless. Nothing is stored between calls. Send the conversation every time.
• From anywhere. The API accepts calls from any origin, so browser tools work. Keys in web pages can be read by visitors: keep them on a server.

AnswersCopy Anchor Link

Answers follow each format's own shape, with three additions:

ErrorsCopy Anchor Link

Errors use each format's own shape, with a stable code to branch on:

{
  "error": {
    "message": "This agent is starting calls faster than 600 a minute. Retry in 1 second.",
    "type": "rate_limit_error",
    "code": "rate_limited",
    "param": null
  }
}

See every code in Errors, and the limits in Rate limits.