Model API

/api/models - Model management endpoints

Overview

The Model API provides AI model configuration and management functionality, supporting custom models and provider-predefined models.

Endpoint List

Model CRUD

Method	Path	Description
GET	/api/models	List models
POST	/api/models	Create custom model
GET	/api/models/:id	Get model details
PATCH	/api/models/:id	Update model
DELETE	/api/models/:id	Delete model

Default Model

Method	Path	Description
GET	/api/models/default	Get default model
PUT	/api/models/default	Set default model

Model Status

Method	Path	Description
POST	/api/models/:id/enable	Enable model
POST	/api/models/:id/disable	Disable model

Provider Models

Method	Path	Description
GET	/api/providers/:id/discover-models	Discover available models
GET	/api/providers/:id/models	List enabled models
POST	/api/providers/:pid/models/:mid/enable	Enable model
POST	/api/providers/:pid/models/:mid/disable	Disable model

---

Detailed Description

GET /api/models

List all models.

Query Parameters:

Parameter	Type	Required	Default	Description
workspace_path	string	No	-	Workspace path
include_global	bool	No	true	Include global models
include_provider_predefined	bool	No	true	Include provider-predefined models

Response:

{
  "models": [
    {
      "id": "claude-3-sonnet",
      "name": "Claude 3 Sonnet",
      "provider": "anthropic",
      "type": "predefined",
      "enabled": true,
      "is_default": true,
      "capabilities": {
        "chat": true,
        "code": true,
        "vision": true,
        "tools": true
      },
      "context_window": 200000,
      "max_output_tokens": 4096
    },
    {
      "id": "my-custom-model",
      "name": "My Custom Model",
      "provider": "openai",
      "type": "custom",
      "enabled": true,
      "base_model": "gpt-4"
    }
  ]
}

---

POST /api/models

Create a custom model.

Request Body:

{
  "id": "my-model",
  "name": "My Custom Model",
  "provider": "openai",
  "base_model": "gpt-4",
  "system_prompt": "You are a helpful assistant.",
  "temperature": 0.7,
  "max_tokens": 4096,
  "parameters": {
    "top_p": 0.9
  }
}

Field	Type	Required	Description
id	string	No	Model ID (auto-generated)
name	string	Yes	Model name
provider	string	Yes	Provider ID
base_model	string	No	Base model
system_prompt	string	No	System prompt
temperature	float	No	Temperature parameter
max_tokens	int	No	Maximum tokens
parameters	object	No	Other parameters

---

GET /api/models/:id

Get model details.

Response:

{
  "id": "claude-3-sonnet",
  "name": "Claude 3 Sonnet",
  "provider": "anthropic",
  "type": "predefined",
  "enabled": true,
  "is_default": true,
  "capabilities": {
    "chat": true,
    "code": true,
    "vision": true,
    "tools": true,
    "streaming": true
  },
  "context_window": 200000,
  "max_output_tokens": 4096,
  "pricing": {
    "input_per_1k": 0.003,
    "output_per_1k": 0.015
  }
}

---

GET /api/providers/:id/discover-models

Discover available models from a provider.

Response:

{
  "models": [
    {
      "id": "claude-3-opus",
      "name": "Claude 3 Opus",
      "enabled": false,
      "description": "Most capable model"
    },
    {
      "id": "claude-3-sonnet",
      "name": "Claude 3 Sonnet",
      "enabled": true,
      "description": "Balanced model"
    },
    {
      "id": "claude-3-haiku",
      "name": "Claude 3 Haiku",
      "enabled": true,
      "description": "Fast and efficient"
    }
  ]
}

---

PUT /api/models/default

Set the default model.

Request Body:

{
  "model_id": "claude-3-sonnet"
}

Response:

{
  "success": true,
  "default_model": "claude-3-sonnet"
}

---

Model Configuration Storage

Model configuration is stored in ~/.viben/models.yaml:

default: claude-3-sonnet

models:
  - id: claude-3-sonnet
    provider: anthropic
    enabled: true

  - id: my-custom-model
    name: My Custom Model
    provider: openai
    base_model: gpt-4
    temperature: 0.7

---

Related Endpoints

• Provider API - Provider management
• Agent API - Agent management