Skip to main content

​
Overview

The Models API provides endpoints to query available AI models configured in DeerFlow. Models are configured in the application’s configuration file and include metadata about their capabilities.

​
List All Models

GET /api/models

Retrieve a list of all available AI models configured in the system

​
Response

​
models
array
List of all configured models with their metadata

​
Example Request

curl http://localhost:8001/api/models

​
Example Response

{
  "models": [
    {
      "name": "gpt-4",
      "display_name": "GPT-4",
      "description": "OpenAI GPT-4 model",
      "supports_thinking": false,
      "supports_reasoning_effort": false
    },
    {
      "name": "claude-3-opus",
      "display_name": "Claude 3 Opus",
      "description": "Anthropic Claude 3 Opus model",
      "supports_thinking": true,
      "supports_reasoning_effort": false
    },
    {
      "name": "o1",
      "display_name": "OpenAI o1",
      "description": "OpenAI o1 reasoning model",
      "supports_thinking": false,
      "supports_reasoning_effort": true
    }
  ]
}

​
Get Model Details

GET /api/models/{model_name}

Retrieve detailed information about a specific AI model by its name

​
Path Parameters

​
model_name
string
required
The unique name of the model to retrieve (e.g., gpt-4, claude-3-opus)

​
Response

​
name
string
required
Unique identifier for the model
​
display_name
string
Human-readable name
​
description
string
Model description
​
supports_thinking
boolean
default:"false"
Whether model supports thinking mode
​
supports_reasoning_effort
boolean
default:"false"
Whether model supports reasoning effort

​
Example Request

curl http://localhost:8001/api/models/gpt-4

​
Example Response

{
  "name": "gpt-4",
  "display_name": "GPT-4",
  "description": "OpenAI GPT-4 model",
  "supports_thinking": false,
  "supports_reasoning_effort": false
}

​
Error Responses

​
404
Not Found
Model not found
{
  "detail": "Model 'invalid-model' not found"
}

​
Model Capabilities

​
Thinking Mode

Models with supports_thinking: true can use extended thinking/reasoning capabilities. This is typically available in Claude models.

​
Reasoning Effort

Models with supports_reasoning_effort: true support configurable reasoning effort levels (e.g., OpenAI’s o1 series). This allows you to control how much computational effort the model applies to reasoning.

​
Use Cases

​
Frontend Model Selection

Use the /api/models endpoint to populate model selection dropdowns: 
const response = await fetch('http://localhost:8001/api/models');
const { models } = await response.json();

// Filter models with specific capabilities
const thinkingModels = models.filter(m => m.supports_thinking);
const reasoningModels = models.filter(m => m.supports_reasoning_effort);

​
Capability Detection

Check if a specific model supports certain features: 
const response = await fetch(`http://localhost:8001/api/models/${modelName}`);
const model = await response.json();

if (model.supports_thinking) {
  // Show thinking mode toggle
}

Gateway Overview

Learn about the Gateway API

Configuration

Configure AI models

Build docs developers (and LLMs) love

Get started for freeTalk to us