Changelog

Major release bringing intelligent multi-region routing and support for 3 new LLM providers.

• FeatureAutomatic multi-region request routing with latency-aware load balancing
• FeatureSupport for Mistral AI, Cohere, and Replicate providers
• ImprovementReduced P99 latency by 28% through improved provider selection algorithm
• FixFixed edge case in fallback provider logic causing occasional timeout errors

Enhanced monitoring and stability improvements.

• FeatureReal-time latency tracking dashboard for each provider
• ImprovementBetter error messages for authentication failures
• FixResolved issue with cache invalidation in concurrent requests
• FixFixed missing X-Request-ID headers in streaming responses

New analytics dashboard and AI-powered cost forecasting capabilities.

• FeatureComprehensive analytics dashboard with cost trends and provider comparison
• FeatureMachine learning-based cost forecasting for next 30-90 days
• FeatureCustom alert thresholds for cost overruns
• ImprovementFaster dashboard load times with data streaming

Enhanced reliability and failover mechanisms.

• ImprovementSmarter failover strategy with exponential backoff
• ImprovementBetter handling of rate-limited providers
• FixFixed stuck requests in failover scenario

Major SDK release with TypeScript support and new streaming capabilities.

• FeatureOfficial TypeScript SDK with full type safety
• FeatureStreaming support in both Python and Node.js SDKs
• ImprovementPython SDK now uses async/await patterns
• BreakingDeprecated synchronous methods in Python SDK

Monitor your costs with real-time alerts and Slack notifications.

• FeatureReal-time cost alerts with configurable thresholds
• FeatureSlack integration for instant notifications
• FeatureEmail digests with daily/weekly summaries
• ImprovementBetter provider health status reporting

Complete platform rewrite with new API architecture and improved performance.

• FeatureRedesigned REST API with better error handling
• FeatureSupport for 15+ new LLM providers
• Improvement50% reduction in API latency
• Breakingv1.x API endpoints are deprecated

InferenceIQ is now in public beta! Join us in optimizing AI inference costs.

• FeaturePublic beta launch with core routing functionality
• FeatureBasic dashboard and usage analytics
• FeatureSupport for OpenAI, Anthropic, and Google Cloud providers
• FeatureFree tier with 1M monthly inference requests