Skip to content

rest_api.md

Latest commit

 

History

History
225 lines (175 loc) · 4.63 KB

rest_api.md

File metadata and controls

225 lines (175 loc) · 4.63 KB

MultiMind REST API

The MultiMind REST API provides a comprehensive interface for using the MultiMind Ensemble system. This API allows you to perform various AI tasks using multiple providers and ensemble methods.

Installation

Make sure you have the MultiMind package installed:

pip install multimind

Starting the API Server

To start the API server:

python -m examples.api.ensemble_api

The server will start on http://localhost:8000. You can access the API documentation at http://localhost:8000/docs.

Available Endpoints

Text Generation

Generate text using an ensemble of models:

POST /generate

Request body:

{
    "prompt": "Your prompt here",
    "providers": ["openai", "anthropic", "ollama"],
    "method": "weighted_voting",
    "weights": {
        "openai": 0.4,
        "anthropic": 0.4,
        "ollama": 0.2
    }
}

Example using Python requests:

import requests

response = requests.post(
    "http://localhost:8000/generate",
    json={
        "prompt": "Explain quantum computing",
        "providers": ["openai", "anthropic", "ollama"],
        "method": "weighted_voting"
    }
)
print(response.json())

Code Review

Review code using an ensemble of models:

POST /review

Request body:

{
    "code": "Your code here",
    "providers": ["openai", "anthropic", "ollama"]
}

Example using Python requests:

import requests

response = requests.post(
    "http://localhost:8000/review",
    json={
        "code": """
        def calculate_factorial(n):
            if n < 0:
                return None
            result = 1
            for i in range(1, n + 1):
                result *= i
            return result
        """,
        "providers": ["openai", "anthropic", "ollama"]
    }
)
print(response.json())

Image Analysis

Analyze images using an ensemble of models:

POST /analyze-image

Request:

  • Form data:
    • image: Image file
  • Query parameters:
    • providers: List of providers (default: ["openai", "anthropic"])

Example using Python requests:

import requests

with open("image.jpg", "rb") as f:
    files = {"image": f}
    response = requests.post(
        "http://localhost:8000/analyze-image",
        files=files,
        params={"providers": ["openai", "anthropic"]}
    )
print(response.json())

Embedding Generation

Generate embeddings using an ensemble of models:

POST /embed

Request body:

{
    "text": "Your text here",
    "providers": ["openai", "huggingface"],
    "weights": {
        "openai": 0.6,
        "huggingface": 0.4
    }
}

Example using Python requests:

import requests

response = requests.post(
    "http://localhost:8000/embed",
    json={
        "text": "This is a sample text",
        "providers": ["openai", "huggingface"]
    }
)
print(response.json())

Response Format

All endpoints return JSON responses with the following structure:

{
    "result": "Generated text or analysis",
    "confidence": 0.95,
    "explanation": "Explanation of the ensemble decision",
    "provider_votes": {
        "provider1": "result1",
        "provider2": "result2",
        ...
    }
}

Error Handling

The API returns appropriate HTTP status codes and error messages:

  • 200 OK: Successful request
  • 400 Bad Request: Invalid request parameters
  • 500 Internal Server Error: Server-side error

Error response format:

{
    "detail": "Error message"
}

Available Ensemble Methods

  • weighted_voting: Combines results based on provider weights
  • confidence_cascade: Uses results based on confidence thresholds
  • parallel_voting: Combines results from all providers in parallel
  • majority_voting: Uses the most common result among providers
  • rank_based: Selects results based on provider ranking

Environment Variables

The API uses the following environment variables (should be set in your .env file):

  • OPENAI_API_KEY: Your OpenAI API key
  • ANTHROPIC_API_KEY: Your Anthropic API key
  • OLLAMA_API_KEY: Your Ollama API key (if using Ollama)
  • HUGGINGFACE_API_KEY: Your Hugging Face API key (if using Hugging Face)

Examples

See the examples/ensemble/usage_examples.py file for comprehensive examples of using the API interface.

API Documentation

The API provides interactive documentation using Swagger UI. You can access it at:

  • Swagger UI: http://localhost:8000/docs
  • ReDoc: http://localhost:8000/redoc

These documentation interfaces allow you to:

  • View all available endpoints
  • Test the API directly from the browser
  • View request/response schemas
  • See example requests and responses