Skip to main content
CC

A local coding agent for your terminal and editor

Codex CLI is OpenAI’s coding agent that runs on your computer from the terminal. It can also be used in VS Code, Cursor, Windsurf, or as a desktop app.

Agenticness = how independently a tool can take action, scored across 9 dimensions. Scored independently by David Kooi, Skylark Creations — see full rubric →

See pricing ↓

Open Source
Code Execution
File Access
B2B
For Developers
CLI
Visit Codex CLI

Is this your tool? Claim this listing to manage your content and analytics.

Recent activity

What's happened with Codex CLI lately

  • Score change
    triggered by news
    Agenticness score 10 → 9 ↓109/36(-1)

    Score re-evaluated after: OpenAI says its Codex workspace agents can run for hours, while Paradigm draws crypto watchers into the mix.

    See the news that prompted this
  • News
    OpenAI says its Codex workspace agents can run for hours, while Paradigm draws crypto watchers into the mix.

    OpenAI launched workspace agents on April 22, 2026, built on Codex and designed for multi-step business workflows that can run for hours. The agents can generate reports, write code, handle Slack interactions, and execute tasks directly in the cloud. For those tracking the AI agent platform ecosystem, this is a major step toward long-running agent execution rather than short chat responses. The surprise is the quiet Paradigm angle, which suggests the release may matter beyond office work and into blockchain security.

    via Crypto Briefing

  • Score change
    Agenticness score 7 → 10 ↑710/36(+3)
  • Score change
    Agenticness score 6 → 7 ↑67/36(+1)

News mentions sourced from our news feed; score changes from periodic re-evaluations.

Ask about Codex CLI

Get answers based on Codex CLI's actual documentation

Try asking:

About

What It Is

Codex CLI is a command-line coding agent from OpenAI for developers who want an AI assistant that works directly in their local development environment. It is designed to run on your computer rather than in a cloud workspace, and the repository is licensed under Apache-2.0.

What to Know

This is a real agentic coding tool, but the public page does not spell out the full level of autonomy or how much confirmation it requires before making changes. The documentation is clearer on setup than on workflow details, so the exact guardrails...

Key Features
Runs locally on your computer
Installed via npm, Homebrew, or downloadable binaries
Supports terminal-based coding workflows
Can be used in VS Code, Cursor, and Windsurf
Offers a desktop app experience
Use Cases
Developers using a terminal-first coding workflow
Editing and iterating on code inside VS Code, Cursor, or Windsurf
Running an AI coding agent locally instead of in a cloud app
Agenticness: Guided Assistant

Executes tasks you assign, one step at a time, within narrow domains.

High evidence
Last evaluated: Jul 9, 2026

Dimension Breakdown

Action Capability
Autonomy
Adaptation
State & Memory
Safety

Categories

Pricing
  • Free: Open-source repository; pricing for use is not fully detailed in the crawled content.
  • Pro / ChatGPT plans: The README says Codex can be used with ChatGPT Plus, Pro, Team, Edu, or Enterprise plans.
  • API key usage: Also supported, but requires additional setup and may involve API costs.
  • Enterprise: Mentioned as part of supported ChatGPT plans.
Details
AddedMarch 26, 2026
RefreshedMarch 28, 2026
Agenticness
Quick Facts
DeploymentOn-device / local
AutonomyCopilot (human-in-loop)
Model supportSingle model
Open sourceYes
Team supportEnterprise
Pricing modelPlatform fee + usage
Interfacecli, gui, ide
Similar tools

Related Tools

Kilo Code is the v5.x community fork of Roo Code (which was archived May 15, 2026). Open-source VS Code extension for agentic coding — Apache 2.0 licensed, BYOK across many providers, runs locally as an IDE extension.

OpenHands (formerly OpenDevin) is a fully autonomous open-source AI coding agent. MIT-licensed core, plans + executes Python actions via CodeAct, uses browser + terminal + files. ~72% on SWE-bench Verified with Sonnet 4.5 via the OpenHands Software Agent SDK (arXiv:2511.03690).

Augment Code is an enterprise-focused AI coding agent built around a 200K-token Context Engine. Closed-source SaaS with strong SWE-bench Verified results (70.6% self-reported) and the highest accuracy on the only public benchmark for AI-assisted code review.

AICode is a VS Code extension for enterprise teams working in large legacy codebases. It maps dependencies and architecture, audits specifications before coding, and keeps changes reviewable inside your infrastructure.

Enterprise
Chrome Extension
B2B
+4
Stay in the loop

Get the weekly agentic AI briefing

New tools, top picks, and trends — delivered every Thursday.

I use AI for: