OpenAI: GPT-5.2-Codex

GPT-5.2-Codex is a paid model from OpenAI that accepts text and image inputs and supports tool use and reasoning. Its context window reaches 400,000 tokens, with up to 128,000 tokens available for completions, making it suitable for long-document workflows, multi-turn conversations, and tasks that require extended output. Structured output support is unconfirmed based on available data. At $1.75 per million input tokens and $14.00 per million output tokens, the output cost is on the higher end, which matters for applications that generate lengthy responses. Its blended benchmark score of 66.4 comes from a single benchmark, so its performance profile is not yet well-validated across diverse tasks. Teams that already rely on OpenAI tooling and need large-context processing with reasoning and vision may find GPT-5.2-Codex worth evaluating, but cost-sensitive users should model their expected output volume carefully before committing.