snailyp/gemini-balance

Features

• API Key Load Balancers - Provides a self-hosted proxy that load balances Gemini API keys with automatic rotation and health checks.
• API Proxy Routings - Serves both Gemini and OpenAI API formats, routing requests to the appropriate backend without client changes.
• OpenAI-Compatible APIs - Accepts requests in OpenAI format and translates them to Gemini API calls for client compatibility.
• Multi-Protocol API Servers - Serves both Gemini and OpenAI API formats, routing requests to the appropriate backend.
• API Key Failure Retriers - Re-attempts failed API calls up to a configurable limit and disables keys that exceed a failure threshold.
• API Key Rotation - Distributes requests across multiple Gemini API keys to avoid rate limits and improve reliability.
• Gemini Key Rotators - Distributes Gemini API requests across multiple keys with automatic rotation and health monitoring.
• Dual-Protocol Endpoints - Accepts requests in both OpenAI and Gemini API formats and routes them to the appropriate backend.
• API Protocol Translations - Translates incoming requests between OpenAI and Gemini API formats at the proxy layer.
• API Key Load Balancers - Distributes requests across multiple API keys with automatic retry and failure detection for reliable service.
• Dual-Format Request Proxies - Accepts chat requests in both Gemini and OpenAI formats and forwards them to the appropriate backend.
• Token Access Restrictions - Checks incoming tokens against a whitelist and rejects unauthorized requests.
• API Key Management - Provides a management interface for adding, monitoring, and rotating multiple Gemini API keys.
• Key Rotation Strategies - Cycles through API keys in round-robin order, distributing requests and disabling failed keys.
• API Key Authentications - Sets Gemini API keys and access tokens through environment variables for authentication and load balancing.
• Request Authentication - Validates incoming requests against a configured token before forwarding them to the upstream API.
• Image Generation Integrations - Converts text prompts into images using Gemini models through an OpenAI-compatible endpoint.
• Text-to-Image Generators - Generates images from text prompts using Gemini models through an OpenAI-compatible endpoint.
• Model Allowlists - Defines which models are enabled, disabled, or reserved for specific features such as image generation.
• Web Search Integrations - Enables models to perform web searches and include result links in responses.
• Database-Backed Configurations - Stores API keys, access tokens, and model settings in a persistent database for live configuration changes.
• Proxy Routing - Overrides the default API base URL with a custom proxy address for Gemini API calls.
• Configuration Hot-Reloading - Changes API keys, model settings, and tokens through a web interface without restarting.
• Live Configuration Reloading - Applies configuration changes instantly through a management interface without requiring a service restart.
• Model - Defines per-model enablement and feature-specific access through configuration.
• Health Dashboards - Tracks API key status and failure counts through a real-time dashboard with automatic recovery.
• Automatic Key Health Testers - Periodically tests disabled API keys and automatically re-enables them upon recovery.
• API Key Status Dashboards - Displays real-time status and failure counts of each API key through an authenticated admin panel.
• Consecutive Failure Detectors - Marks keys as invalid only after a configurable number of consecutive failures to avoid false positives.
• Web Management Interfaces - Provides a browser-based interface for managing API keys and configuration settings.
• Model Selection Filters - Removes unwanted models from the available list through simple configuration settings.
• Persistent Token Authenticators - Uses a database-persisted token for management interface access, ignoring later environment variable changes.
• Key Recovery - Periodically tests disabled API keys and re-enables them when they pass consecutive health checks.
• Real-Time Monitoring Dashboards - Ships a real-time dashboard for monitoring the status and usage of each API key.
• Database-Backed Token Validators - Validates every incoming request against a pre-configured token whitelist stored in the database.

Star-Verlauf