One API key. 23+ models. Switch with a single line of code.
22 models
Most capable multimodal model with vision, images, and audio support
Best-in-class reasoning, coding, and complex analysis
Advanced chain-of-thought reasoning for hard problems
Fast reasoning for code and quick tasks
Open-source deep reasoning model, extremely capable
Cost-efficient frontier model with excellent quality
128K context, most powerful GPT-4 variant
Latest Claude flagship with enhanced reasoning
Fast and affordable small model for simple tasks
Lightning fast, low cost, great for high-volume tasks
Google's fastest flagship, 1M token context
Production-ready fast model with excellent quality
128K context optimized for RAG and retrieval
Extended 32K context window
Maximum capability, 200K context window
State-of-the-art speech recognition, 98+ languages
Natural human-like voice synthesis, 50+ voices
1536 dimensions, best quality embeddings
Efficient and affordable embeddings
Latest image generation with photorealistic quality
Reliable fast chat model, battle-tested
Original GPT-4, still capable for many tasks
