# X6.AI | OpenAPI Documentation

> OpenAI-compatible API. Endpoint (no /v1): https://api.x6.ai

## Instructions for AI Agents

- Append `.md` to any doc page URL for clean Markdown (e.g. `/en/docs/api/quickstart.md`)
- Index: https://x6.ai/api/llms.txt
- Full reference: https://x6.ai/api/llms-full-en.txt
- OpenAPI spec: https://x6.ai/api/openapi.yaml
- Endpoint host: https://api.x6.ai — do not put /v1 on the host itself
- OpenAI SDK: base_url=https://api.x6.ai/v1 — host + /v1
- Anthropic SDK: base_url=https://api.x6.ai — host only, no /v1

## Getting Started
- [Endpoints](https://x6.ai/en/docs/api/endpoints.md): Single API base URL
- [Quickstart](https://x6.ai/en/docs/api/quickstart.md): First chat completion
- [Authentication](https://x6.ai/en/docs/api/authentication.md): Bearer API key format and errors

## API Reference
- [Authentication](https://x6.ai/en/docs/api/authentication.md): Bearer API key format and errors
- [Chat Completions](https://x6.ai/en/docs/api/chat-completions.md): POST /v1/chat/completions reference
- [Messages (Anthropic)](https://x6.ai/en/docs/api/messages.md): POST /v1/messages Anthropic-compatible API
- [Models](https://x6.ai/en/docs/api/models.md): GET /v1/models list and retrieve
- [Model pricing](https://x6.ai/en/docs/api/model-pricing.md): Where to find input / cached / output rates
- [Streaming](https://x6.ai/en/docs/api/streaming.md): SSE streaming responses
- [Parameters](https://x6.ai/en/docs/api/parameters.md): Request parameters
- [Errors](https://x6.ai/en/docs/api/errors.md): HTTP status codes and error shapes
- [Rate Limits & Quotas](https://x6.ai/en/docs/api/rate-limits-and-quotas.md): IP limits, quota and balance pre-checks
- [Response Headers](https://x6.ai/en/docs/api/response-headers.md): X-SNX-* diagnostic headers

## Quick start integrations
- [OpenAI SDK Integration](https://x6.ai/en/guides/quick-start/openai-sdk-integration) · [MD](https://x6.ai/en/docs/api/openai-sdk-integration.md): Drop-in OpenAI SDK migration
- [Anthropic SDK Integration](https://x6.ai/en/guides/quick-start/anthropic-sdk-integration) · [MD](https://x6.ai/en/docs/api/anthropic-sdk-integration.md): Anthropic SDK and Claude Code via /v1/messages
- [DeepSeek Harness](https://x6.ai/en/guides/quick-start/deepseek-harness) · [MD](https://x6.ai/en/docs/api/deepseek-harness.md): Custom provider and recommended API for DeepSeek Harness

## Optional
- [Developer Console](https://console.x6.ai): API keys, projects, usage
- [OpenAPI JSON](https://x6.ai/api/openapi.json): machine-readable schema