OpenAI: GPT-5 Codex

GPT-5 Codex is a paid model from OpenAI that accepts text and image inputs and supports both tool use and reasoning. Its 400,000-token context window accommodates long documents or multi-turn sessions, and the 128,000-token output ceiling suits tasks that require extended completions. Structured output support is not confirmed in current specifications, so workflows that depend on guaranteed JSON schemas should be verified before committing. At $1.25 per million input tokens and $10.00 per million output tokens, output costs are on the higher end of the market, making volume a real consideration. The blended benchmark score of 59.6 comes from a single independent benchmark, so treat that figure as a preliminary signal rather than a settled verdict. Teams that need long-context multimodal reasoning with tool integration are the natural fit, but budget-sensitive users should weigh the output pricing carefully against cheaper alternatives with broader benchmark coverage.