Available Models
GPT-4o OpenAI’s versatile multimodal model. 2x faster and 50% cheaper than GPT-4 Turbo while matching intelligence. Excellent for vision tasks, non-English languages, and general-purpose use.
GPT-4.1 OpenAI’s flagship model for coding and instruction-following. Achieves 54.6% on SWE-bench Verified, 87.4% on IFEval, with 1M token context window. Optimized for complex agentic workflows and tool use.
GPT-4o Mini Cost-effective multimodal model for high-volume applications. Fast responses with strong reasoning. Best value for chat, summarization, and classification.
GPT-5 OpenAI’s most advanced reasoning model. Dynamic reasoning with expert-level responses, ~80% fewer hallucinations. Unified architecture combining GPT-4 and o-series capabilities.
GPT-4.1 Mini Balanced model competitive with GPT-4o at lower cost. 1M context, 45.1% on hard instruction evals. Ideal for coding, RAG, and agentic applications.
OpenAI o1 OpenAI’s first reasoning model with chain-of-thought. Excels at STEM, coding, and complex problem-solving with adjustable reasoning effort.
OpenAI o3 OpenAI’s most capable reasoning model. Sets new standards for math, science, coding, and visual reasoning. Excels at technical writing and complex instructions.
GPT-4.1 Nano Smallest model in GPT-4.1 family with exceptional efficiency. 80.1% MMLU, 50.3% GPQA. 1M context at the lowest cost for embeddings and lightweight tasks.
Sora 2 Sora 2 from OpenAI generates photorealistic videos up to 20 seconds with exceptional understanding of physics, motion, and spatial relationships. Creates coherent scenes with complex camera movements.