Codex Mini

Codex Mini is a compact language model exposed on OpenRouter under the OpenAI namespace, identified by the model ID openai/codex-mini. According to the OpenRouter listing documentation, it advertises a 200,000-token context window and supports both image and text inputs, with parameters exposed for reasoning, structured outputs, seed, and function-calling tools. The model is positioned as a lightweight variant within the Codex family, intended for coding and general text tasks where lower latency and cost matter more than the deepest reasoning performance.

Beyond its reasoning and tool-use parameters, Codex Mini's practical fit centers on developer workflows that benefit from a small, instruction-following model with function-calling and structured-output capabilities. The OpenRouter route surfaces support for response format controls and tool choice, making it usable for agent-style pipelines that call external functions or require deterministic JSON outputs. Its combination of a large context window with vision-capable input gives it flexibility for tasks involving codebases paired with screenshots or diagrams, while its compact footprint makes it suitable for cost-sensitive or high-throughput applications.