looplj/axonhub
Axonhub
Axonhub is an AI gateway and multi-model API proxy that provides a unified interface for routing requests to multiple large language model providers. It functions as a load balancer and translation layer, converting a standardized API format into provider-specific payloads to enable communication with various AI models without provider-specific code.
The system manages traffic through rule-based routing and automatic failover to maintain high availability. It differentiates its operations by providing a provider-agnostic interface that decouples client requests from specific model backends using abstract identifiers.
The platform incorporates AI cost monitoring to calculate per-request spending based on token usage and an observability platform for thread-aware request tracing. Security is handled through role-based access control, usage quotas, and restricted API key permissions to ensure data isolation.
The gateway supports the generation of multiple content types, including text, images, vector embeddings, and reranked results.
Features
- LLM Gateways - Provides a centralized interface that aggregates multiple LLM providers into a single standardized API.
- AI Model Load Balancers - Functions as an AI model load balancer that distributes requests across providers with automatic failover.
- AI Access Control - Applies fine-grained authorization logic and usage quotas to secure AI model access via API keys.
- API Format Standardizers - Translates a single unified request format into provider-specific payloads for multiple LLM vendors.
- API Request Translations - Converts a standardized interface request into provider-specific API parameters for multiple AI models.
- Model Routing - Maps abstract model identifiers to specific provider channels to ensure automatic failover and operational optimization.
- Provider-Agnostic Model Interfaces - Provides an abstraction layer that standardizes inputs and outputs across multiple LLM providers via abstract identifiers.
- Unified AI API Endpoints - Provides a single standardized API endpoint that translates requests across multiple AI provider formats.
- Traffic Routing Rules - Uses rule-based routing and failover logic to direct traffic to specific model channels and maintain availability.
- AI Observability Tracing - Provides thread-aware observability to analyze the complete timeline of AI requests and identify interaction failures.
- Multi-Modal Provider Proxies - Acts as a proxy layer that translates and unifies requests for text, images, and embeddings across different providers.
- API Key Scopes - Restricts API keys to specific model mappings and profiles to enforce the principle of least privilege.
- Role-Based Access Controls - Enforces usage limits, role-based permissions, and data isolation to secure AI resources for different user groups.
- Role-Based Access Control - Implements role-based access control to manage user permissions and restrict access to specific model mappings.
- AI Cost Monitoring - Tracks token usage and calculates per-request spending to monitor and optimize operational expenses.
- End-to-End Message Tracing - Captures complete execution timelines and metadata for AI interactions to provide end-to-end request observability.
- Token Cost Calculators - Calculates real-time spending by applying provider pricing to input and output token counts.
Star history
How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.
