Unified API Access
Access over 300 AI models from more than 60 providers through one API endpoint compatible with the OpenAI SDK.
Automatic Fallbacks and Routing
Ensures reliability by automatically routing requests to alternative providers if one fails and optimizes costs by selecting preferred or cheaper options.
Prompt Caching
Caches prompts to reduce usage costs across supported models.
Data Policy-Based Routing
Allows control over which models and providers receive specific prompts based on data policies.
DevTools for SDK Telemetry
Provides tools to capture requests, responses, token usage, and errors during development (currently pre-release).