DeepSeek
Table of contents
Use DeepSeek models with Docker Agent.
Overview
DeepSeek serves its frontier chat and reasoning models through an OpenAI-compatible API, with strong price/performance on coding and reasoning tasks. Docker Agent includes built-in support for DeepSeek as an alias provider.
Setup
Create an API key from the DeepSeek Platform.
Set the environment variable:
export DEEPSEEK_API_KEY=your-api-key
Usage
Inline Syntax
The simplest way to use DeepSeek:
agents:
root:
model: deepseek/deepseek-v4-pro
description: Assistant using DeepSeek
instruction: You are a helpful assistant.Named Model
For more control over parameters:
models:
deepseek_model:
provider: deepseek
model: deepseek-v4-pro
temperature: 0.7
max_tokens: 8192
agents:
root:
model: deepseek_model
description: Assistant using DeepSeek
instruction: You are a helpful assistant.Available Models
DeepSeek exposes a small, vendor-controlled model lineup. Check the DeepSeek models documentation for current model IDs, context limits, and pricing.
| Model | Description |
|---|---|
deepseek-v4-pro | DeepSeek V4 Pro, extended-reasoning model with tool calling |
deepseek-v4-flash | DeepSeek V4 Flash, fast general-purpose model with tool calling |
deepseek-v4-flash-vision-exp | DeepSeek V4 Flash Vision (experimental), multimodal model with image input |
Model IDs are case-sensitive and must be passed exactly as the catalogue lists them.
How It Works
DeepSeek is implemented as a built-in alias in Docker Agent:
- API Type: OpenAI-compatible (
openai_chatcompletions) - Base URL:
https://api.deepseek.com/v1 - Token Variable:
DEEPSEEK_API_KEY
Example: Code Assistant
agents:
coder:
model: deepseek/deepseek-v4-pro
description: Code assistant using DeepSeek V4 Pro
instruction: |
You are an expert programmer.
Write clean, well-documented code and follow language best practices.
toolsets:
- type: filesystem
- type: shell
- type: think