OpenAI: o4 Mini

OpenAI: o4 Mini is a reasoning-capable model from OpenAI that accepts text, images, and files as input, with a 200,000-token context window and up to 100,000 tokens of output per completion. It supports tool use and reasoning, making it applicable to multi-step workflows and document-heavy tasks. Structured output support is unconfirmed from available data. At $1.1 per million input tokens and $4.4 per million output tokens, it sits in a mid-range price tier. Its blended benchmark score of 64.5 covers only three benchmarks, so that figure should be treated as provisional rather than comprehensive. Teams doing code-heavy work may find the aider_polyglot score of 72.0 relevant, but buyers requiring broad, independently verified performance data should weigh that limited coverage before committing.