See what we're building next. Our public roadmap keeps you informed about upcoming features, improvements, and new models.
Our next-generation foundation model with 2x performance, improved reasoning, and 500K context window.
Enhanced function calling with parallel execution, streaming, and improved reliability.
Detailed usage analytics, cost breakdowns, and performance insights directly in your dashboard.
Advanced multimodal capabilities with video understanding, document analysis, and image generation.
Single sign-on with SAML/OIDC and automated user provisioning for enterprise customers.
Custom model fine-tuning with your own data for specialized use cases.
Deploy models to edge locations worldwide for sub-10ms latency responses.
Build and deploy autonomous AI agents with memory, planning, and tool use capabilities.
New models and model improvements
API features and enhancements
Enterprise features and security
SDKs, libraries, and tooling
Real-time token streaming for all models with improved latency.
Completely rewritten Python SDK with async support and type hints.
Set spending limits and receive alerts when approaching thresholds.
Have an idea? We'd love to hear from you. Submit your feature request and vote for others.