> ## Documentation Index
> Fetch the complete documentation index at: https://docs.openpawz.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# DeepSeek

> Use DeepSeek's reasoning and coding models.

# DeepSeek

Use DeepSeek's reasoning and coding models.

## Setup

1. Get an API key at [platform.deepseek.com](https://platform.deepseek.com/)
2. In Pawz → **Settings → Providers** → **Add Provider** → **DeepSeek**
3. Paste your API key

## Configuration

| Setting | Default |
| - | - |
| Base URL | `https://api.deepseek.com/v1` |
| API key | Required |
| Default model | — |

## Models

| Model | Context | Best for |
| - | - | - |
| `deepseek-chat` | 64K | General conversation, coding, analysis |
| `deepseek-reasoner` | 64K | Complex multi-step reasoning with chain-of-thought |

## Reasoning tokens

The `deepseek-reasoner` model uses extended "thinking" tokens during complex tasks. These reasoning tokens count toward your usage but produce significantly better results on math, logic, and multi-step problems. Pawz displays thinking content in a collapsible block in the chat UI.

## Cost

DeepSeek is one of the most affordable providers — significantly cheaper than OpenAI or Anthropic for comparable quality. A great choice as a `worker_model` in multi-agent orchestration.

## Tips

:::tip

* **Use `deepseek-chat` as your `cheap_model`** — excellent quality at a fraction of frontier model pricing.
* **Pair with Ollama** — test locally with Ollama, then use DeepSeek for production tasks that need a hosted model.
* DeepSeek uses an OpenAI-compatible API, so it works through the standard `OpenAiProvider`.
  :::


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.