173 models across text, image, audio & embedding
Latest GPT flagship model with 1M context window. Extended thinking, multimodal input, and document editing capabilities. Outperforms GPT-5.3 Codex on all benchmarks.