List All Models

MakeHub provides access to a variety of AI models from different providers. This page explains how to retrieve information about the available models and how to interpret the response data.

Listing Available Models

You can retrieve a list of all available models through the /v1/models endpoint:

GET https://api.makehub.ai/v1/models

This endpoint requires authentication with your MakeHub API key.

Python Example

import requests
 
API_BASE_URL = "https://api.makehub.ai"
API_KEY = "your_makehub_api_key"
 
def list_models():
    headers = {
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    }
    
    response = requests.get(f"{API_BASE_URL}/v1/models", headers=headers)
    
    if response.status_code == 200:
        return response.json()
    else:
        return {"error": response.text}
 
models = list_models()
print(models)

TypeScript Example

import axios from 'axios';
 
async function listModels() {
  try {
    const response = await axios.get('https://api.makehub.ai/v1/models', {
      headers: {
        'Authorization': `Bearer your_makehub_api_key`,
        'Content-Type': 'application/json'
      }
    });
    
    return response.data;
  } catch (error) {
    console.error('Error fetching models:', error);
    return { error: error.message };
  }
}
 
// Example usage
listModels().then(data => console.log(data));

cURL Example

curl "https://api.makehub.ai/v1/models" \
  -H "Authorization: Bearer your_makehub_api_key" \
  -H "Content-Type: application/json"

Response Format

The response contains an array of model objects under the data key. Each model object includes the following properties:

Property	Type	Description
`model_id`	string	Unique identifier for the model (in the format `organisation/model_name`)
`model_name`	string	Name of the model
`organisation`	string	Organisation that created the model
`provider_name`	string	Service provider hosting the model
`context`	integer	Maximum context length in thousands of tokens
`price_per_input_token`	number	Cost per input token in cents/1M tokens
`price_per_output_token`	number	Cost per output token in cents/1M tokens
`quantisation`	string or null	Quantisation level, if applicable
`native_support_function_calling`	boolean	Whether the model natively supports function calling
`native_support_tools_calling`	boolean	Whether the model natively supports tools calling

Sample Response

{
  "data": [
    {
      "context": 200,
      "model_id": "anthropic/claude-3-5-haiku",
      "model_name": "claude-3-5-haiku",
      "organisation": "anthropic",
      "price_per_input_token": 0.8,
      "price_per_output_token": 4,
      "provider_name": "anthropic",
      "quantisation": null,
      "native_support_function_calling": true,
      "native_support_tools_calling": true
    },
    {
      "context": 200,
      "model_id": "anthropic/claude-3-5-sonnet",
      "model_name": "claude-3-5-sonnet",
      "organisation": "anthropic",
      "price_per_input_token": 3,
      "price_per_output_token": 15,
      "provider_name": "anthropic",
      "quantisation": null,
      "native_support_function_calling": true,
      "native_support_tools_calling": true
    },
    {
      "context": 131,
      "model_id": "openai/gpt-4o",
      "model_name": "gpt-4o",
      "organisation": "openai",
      "price_per_input_token": 2.5,
      "price_per_output_token": 10,
      "provider_name": "azure-UKSouth",
      "quantisation": null,
      "native_support_function_calling": true,
      "native_support_tools_calling": true
    }
    // Additional models omitted for brevity
  ]
}

Finding the Most Cost-Effective Model

def find_cheapest_model_for_task(min_context_size=32, require_function_calling=False):
    models_data = list_models()
    
    # Filter models based on requirements
    eligible_models = [
        model for model in models_data["data"] 
        if model.get("context", 0) >= min_context_size and
        (not require_function_calling or model.get("native_support_function_calling", False))
    ]
    
    # Sort by output token price (most common cost driver)
    eligible_models.sort(key=lambda x: x.get("price_per_output_token", float('inf')))
    
    return eligible_models[0] if eligible_models else None

Real-time Metrics Native Tools Calling