Skip to main content
Model types define the different AI/ML providers supported by Syft Space. Each type has its own configuration requirements and capabilities.

​
List model types

Get all available model types with their configuration schemas.

​
Response

​
name
string
Model type identifier used in the dtype field when creating models.
​
description
string
Human-readable description of the model type and its capabilities.
​
icon
string
Icon representation for the model type.
​
enabled
boolean
Whether this model type is currently enabled. Disabled types may lack required dependencies.
​
config_schema
object
JSON schema describing the required and optional configuration fields for this model type.
curl https://your-server.com/api/v1/models/types \
  -H "Authorization: Bearer YOUR_API_KEY"

[
  {
    "name": "openai",
    "description": "OpenAI model type for interacting with OpenAI's API.\n\n    Supports OpenAI's chat completion API and compatible endpoints.\n    Can be configured with API key, model selection, and custom base URL.\n\n    Reference: https://platform.openai.com/docs/api-reference",
    "icon": "🤖",
    "enabled": true,
    "config_schema": {
      "type": "object",
      "properties": {
        "api_key": {
          "type": "string",
          "title": "OpenAI API Key",
          "description": "Your OpenAI API key"
        },
        "model": {
          "type": "string",
          "title": "Model",
          "description": "Model to use for chat completions",
          "default": "gpt-3.5-turbo"
        },
        "base_url": {
          "type": "string",
          "title": "Base URL",
          "description": "Custom base URL for OpenAI-compatible APIs (optional)",
          "default": "https://api.openai.com/v1"
        },
        "system_prompt": {
          "type": "string",
          "title": "System Prompt",
          "description": "System prompt to use for chat completions",
          "default": ""
        }
      },
      "required": ["api_key"],
      "order": ["api_key", "model", "base_url"]
    }
  }
]

​
Endpoint

GET

​
Get model type

Get detailed information about a specific model type.

​
Path parameters

​
name
string
required
Model type identifier (e.g., openai, anthropic).

​
Response

Returns a single model type object with the same structure as the list endpoint. 
curl https://your-server.com/api/v1/models/types/openai \
  -H "Authorization: Bearer YOUR_API_KEY"

{
  "name": "openai",
  "description": "OpenAI model type for interacting with OpenAI's API.\n\n    Supports OpenAI's chat completion API and compatible endpoints.\n    Can be configured with API key, model selection, and custom base URL.\n\n    Reference: https://platform.openai.com/docs/api-reference",
  "icon": "🤖",
  "enabled": true,
  "config_schema": {
    "type": "object",
    "properties": {
      "api_key": {
        "type": "string",
        "title": "OpenAI API Key",
        "description": "Your OpenAI API key"
      },
      "model": {
        "type": "string",
        "title": "Model",
        "description": "Model to use for chat completions",
        "default": "gpt-3.5-turbo"
      },
      "base_url": {
        "type": "string",
        "title": "Base URL",
        "description": "Custom base URL for OpenAI-compatible APIs (optional)",
        "default": "https://api.openai.com/v1"
      },
      "system_prompt": {
        "type": "string",
        "title": "System Prompt",
        "description": "System prompt to use for chat completions",
        "default": ""
      }
    },
    "required": ["api_key"],
    "order": ["api_key", "model", "base_url"]
  }
}

​
Endpoint

GET

​
Get model type schema

Get only the configuration schema for a specific model type.

​
Path parameters

​
name
string
required
Model type identifier.

​
Response

Returns the JSON schema object for the model type’s configuration. 
curl https://your-server.com/api/v1/models/types/openai/schema \
  -H "Authorization: Bearer YOUR_API_KEY"

{
  "type": "object",
  "properties": {
    "api_key": {
      "type": "string",
      "title": "OpenAI API Key",
      "description": "Your OpenAI API key"
    },
    "model": {
      "type": "string",
      "title": "Model",
      "description": "Model to use for chat completions",
      "default": "gpt-3.5-turbo"
    },
    "base_url": {
      "type": "string",
      "title": "Base URL",
      "description": "Custom base URL for OpenAI-compatible APIs (optional)",
      "default": "https://api.openai.com/v1"
    },
    "system_prompt": {
      "type": "string",
      "title": "System Prompt",
      "description": "System prompt to use for chat completions",
      "default": ""
    }
  },
  "required": ["api_key"],
  "order": ["api_key", "model", "base_url"]
}

​
Endpoint

GET

​
Supported model types

Syft Space supports the following model types:

​
OpenAI

Type identifier: openai Supports OpenAI’s chat completion API and OpenAI-compatible APIs (e.g., Azure OpenAI, LocalAI, LM Studio). Configuration:
  • api_key (required): OpenAI API key
  • model: Model identifier (default: gpt-3.5-turbo)
  • base_url: Custom API endpoint (default: https://api.openai.com/v1)
  • system_prompt: Default system message
Example models: gpt-4, gpt-4-turbo, gpt-3.5-turbo OpenAI-compatible providers: The OpenAI model type supports any provider with an OpenAI-compatible API. Use the base_url parameter to connect to alternative providers: 
{
  "dtype": "openai",
  "configuration": {
    "base_url": "http://localhost:11434/v1",
    "api_key": "not-needed",
    "model": "llama2"
  }
}
Supports locally-hosted Ollama models. Available models depend on what you’ve pulled (e.g., llama2, mistral, codellama, phi).
Currently, only the OpenAI model type is implemented. Support for native Anthropic API (non-OpenAI-compatible) is planned for future releases.

Build docs developers (and LLMs) love

Get started for freeTalk to us