Back to codex
OpenAI /

AI
codex-mini-latest

Closed SourceCodingtextUpdated July 14, 2026

Codex Mini

Model Overview

Codex Mini (slug: codex-mini) is a series of lightweight, cost-efficient models developed by OpenAI, optimized for rapid, agentic coding tasks within the Codex product ecosystem (such as the Codex CLI and VS Code extensions). Often evolving as a series rather than a static model (e.g., GPT-5-Codex or GPT-5.1-Codex-Mini), it is deployed as an API-only, closed-source model that prioritizes low latency and high throughput for coding workflows.

Capabilities

  • Code Generation: Excels at CLI workflows, code editing, and refactoring.
  • Agentic Tasks: Optimized for automating shell commands and acting as a quick-response subagent for routine coding work.
  • Efficiency: Designed for rapid response times, making it ideal for immediate, low-overhead interactions during development.

Example Use Cases

  • Automating routine terminal commands via the Codex CLI.
  • Providing fast, inline code completions and refactoring suggestions in IDEs.
  • Routing straightforward coding tasks to a cost-effective model, reserving larger models (like GPT-4o or GPT-5) for complex, repository-wide reasoning.

Performance & Benchmarks

Due to the rapidly evolving nature of the "codex-mini" series (where the underlying model is frequently updated), official static benchmark scores are rarely emphasized. Performance is typically measured by developer productivity—such as the number of coding tasks completed within a specific latency and cost budget. It sacrifices peak reasoning capability for speed and cost-efficiency.

Intended Use & Limitations

  • Intended Use: Designed for integration into developer tools where low latency and cost-efficiency are paramount for simple to moderate coding tasks.
  • Limitations: Not intended for complex, architecture-level reasoning or large-scale repository refactoring, which are better suited for OpenAI's frontier models.

About OpenAI

OpenAI is an artificial intelligence research laboratory and deployment company dedicated to ensuring that artificial general intelligence (AGI) benefits all of humanity. They are the creators of the GPT series, Codex, and ChatGPT.

You might also want to compare

Verified Sources

Model Specs

closed-source

Parameters

Unknown

Context Window

Unknown

License

Proprietary

Deployment

api-only

Resources & Links

Lineage

Model Family

Part of the codex family

Only release in this line currently tracked.

Curator Notes

Bulk imported from OpenAI developer docs.

Compare Specs

Compare parameters, context windows, modalities, and benchmark scores of this model side-by-side with others.

Compare Model