# Private Model Gateway

> Private OpenAI-compatible chat completions API at https://llm.hisavage.com/v1 (Bearer key auth, neutral model names default/smart/fast plus pinned versions).

## Docs

- [Full documentation (markdown)](https://llm.hisavage.com/docs.md): everything an agent needs: auth, endpoints, model table, examples, limits, errors
- [Full documentation inline](https://llm.hisavage.com/llms-full.txt)
- [HTML docs](https://llm.hisavage.com/docs)

## Sections

- [Instructions for AI agents](https://llm.hisavage.com/docs#instructions-for-ai-agents)
- [Endpoints](https://llm.hisavage.com/docs#endpoints)
- [Models](https://llm.hisavage.com/docs#models)
- [Examples](https://llm.hisavage.com/docs#examples)
- [Vision / images](https://llm.hisavage.com/docs#vision-images)
- [Tool calling](https://llm.hisavage.com/docs#tool-calling)
- [JSON / structured output](https://llm.hisavage.com/docs#json-structured-output)
- [Limits](https://llm.hisavage.com/docs#limits)
- [Errors and retries](https://llm.hisavage.com/docs#errors-and-retries)
- [Quick self-test](https://llm.hisavage.com/docs#quick-self-test)