justlovemaki/AIClient-2-API
AIClient 2 API
AIClient-2-API is an AI gateway and proxy server that translates diverse large language model interfaces into a single standardized API format. It functions as an OpenAI API compatible proxy and multi-provider orchestrator, allowing a single client to interact with multiple different model backends through a unified interface.
The project distinguishes itself by acting as a load balancer that distributes requests across multiple provider accounts using health checks and polling to bypass quota limits. It includes a TLS fingerprint emulator to simulate browser characteristics and prevent API access blocks, as well as a failover system that automatically redirects requests to secondary providers when a primary service is unhealthy.
The system covers broad capability areas including traffic management via network proxy configuration, model behavior control through system prompt overrides, and the processing of multimodal inputs such as images and documents. It also provides API traffic logging for auditing and debugging.
The server is packaged as a Docker image for containerized deployment.
Features
- AI Gateways - Functions as a unified gateway for managing and routing requests across multiple AI model providers.
- LLM Gateways - Functions as a centralized gateway that aggregates multiple LLM providers into a single standardized API.
- OpenAI-Compatible APIs - Implements HTTP endpoints that provide a standardized OpenAI-compatible format for interacting with diverse machine learning models.
- AI Provider Integrations - Connects multiple third-party model providers to a single client through a unified interface.
- API Format Standardizers - Translates diverse language model interfaces into a single compatible format for use with existing AI tools and agents.
- Model API Integrations - Provides connectors that allow software to interact with various large language model services via standardized protocols.
- Model Provider Interfaces - Provides a standardized layer for normalizing responses from various artificial intelligence service providers into a single interface.
- Model Fallbacks - Automatically switches to alternative AI models or providers when the primary service is unhealthy or returns an error.
- Provider-Agnostic Request Normalization - Converts diverse request formats from different model providers into a standardized interface to ensure SDK compatibility.
- Quota Routing - Distributes requests across multiple provider accounts using polling and health checks to bypass API quota limits.
- AI Model Load Balancers - Provides a load balancer optimized for distributing requests across multiple AI model providers using health checks and polling.
- Dual-Format Request Proxies - Acts as a proxy that accepts multiple API formats and translates them into the appropriate backend provider protocols.
- AI Provider Managers - Provides a management console for modifying provider parameters, monitoring health, and managing API credentials.
- System Prompt Overrides - Implements a middleware system to intercept and override system prompts to standardize AI responses across different providers.
- Runtime Provider Switching - Enables switching the active backend AI provider at runtime via configuration without modifying the client application.
- Multimodal Input Processors - Processes diverse input modalities including images and documents for consumption by supported large model providers.
- System Prompts - Injects structured system instructions into models to enforce specific behaviors and operational constraints across different providers.
- Runtime Behavior Configuration - Allows tuning of proxy behavior and model interaction parameters through a central runtime configuration file.
- AI Server Containerization - Packages the AI proxy server and its API interfaces into a Docker container for simplified hosting.
- Containerized Deployments - Packages the translation layer into a Docker image to ensure consistent execution across different hosting platforms.
- Dynamic Configuration Providers - Allows adjusting service behavior and provider settings at runtime via configuration without requiring server restarts.
- TLS Fingerprint Emulators - Simulates browser-specific TLS handshake parameters to bypass security filters that block non-browser client requests.
- Request Interception Middleware - Intercepts request payloads to append or replace system instructions before forwarding them to the model provider.
