Changelog
Keep up with the latest updates, improvements, and fixes.
Semantic Caching GA
Semantic caching is now generally available. Reduce costs by up to 80% on repeated similar queries with automatic cache invalidation.
- Automatic semantic similarity detection
- Configurable cache TTL per model
- Cache analytics in dashboard
Regional Model Filtering
Filter and route requests to models deployed in specific regions for data sovereignty and compliance requirements.
- US, EU, and APAC region support
- Per-request region override
- Compliance reporting dashboard
Improved Fallback Latency
Reduced auto-fallback detection time from 5s to 500ms with health-check-based routing.
- 500ms fallback detection
- Proactive health monitoring
- Fallback chain visualization
BYOK Support
Bring Your Own Key — use your existing provider API keys through AnyRouter for direct billing with full routing benefits.
- Support for OpenAI, Anthropic, Google keys
- Automatic key rotation
- Split billing reports
Audit Logging with R2
Complete request/response audit trail stored in Cloudflare R2 with configurable retention policies.
- Full request/response logging
- Configurable retention (7-365 days)
- Export to S3-compatible storage
Rate Limit Fix
Fixed an issue where rate limits were not correctly applied to BYOK requests.
- Correct rate limiting for BYOK
- Improved error messages
AnyRouter Launch
Initial release of AnyRouter — the edge-native AI gateway. Route to 150+ models from 28 providers with a single API.
- 150+ AI models supported
- 28 provider integrations
- OpenAI-compatible API
- Auto-fallback routing
- Usage analytics dashboard