> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.mixlayer.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.mixlayer.com/_mcp/server.

# Models

> Model families available through Mixlayer

Mixlayer hosts a curated set of open-source models behind its OpenAI-compatible inference APIs.

## Model families

| Family                                                                                                                       | Best for                                                                            |
| ---------------------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------- |
| [Qwen 3.5/3.6](/qwen-3-5) A range of model sizes with vision, tool calling, and optional thinking.                           | General-purpose applications, from free prototyping to demanding agentic workloads. |
| [Kimi K2.7 Code](/kimi-k2-7-code) A coding-specialized model with always-on thinking.                                        | Long-horizon software engineering, repository work, and coding agents.              |
| [GLM 5.2](/glm-5-2) A flagship model designed for long-context, long-horizon tasks.                                          | Large codebases, research, complex tool use, and sustained agent workflows.         |
| [GLM 5.3](/glm-5-3) Always-on reasoning with native Low, High, and Max effort levels.                                        | Text and tool-calling workflows that need control over reasoning effort.            |
| [DeepSeek V4.1 Flash](/deepseek-v4-1-flash) A text model with optional thinking and native Low, High, and Max effort levels. | Coding agents, large document collections, and reasoning tasks.                     |

> **Info**
>
> Exact model identifiers, context limits, capabilities, and availability is
> available in the Mixlayer console or from the `GET /v1/models` API endpoint.
> Pricing is listed on the [Pricing](/pricing) page.

## Use a model

Pass a model identifier as the `model` field:

```json
{
  "model": "model-identifier",
  "messages": [{"role": "user", "content": "Hello!"}]
}
```

Choose the smallest model that meets your quality requirements. Evaluate
candidates on representative prompts before selecting one for production.

## Don't see a model that works for you?

Mixlayer can host custom, fine-tuned, and distilled models for you.
[Get in touch](mailto:hello@mixlayer.com) to learn more.