OpenAI Codex Verified
Key insights
Concrete technical or product signals.
- GPT-6.1 Sol's Codex rollout includes Plus, Pro, Business, Enterprise, and Edu in the desktop app and CLI, plus ChatGPT Work on the web and mobile. Enterprise and Edu keep it off until an administrator enables it. Free and Go are not included at launch.
- GPT-5.5 retires from ChatGPT, ChatGPT Work, and Codex on October 14, 2026. That retirement does not apply to the OpenAI API. The models page tells Plus and higher plans to choose GPT-6 Sol when available, and Free and Go to choose GPT-6 Luna in the desktop app.
- Codex Cloud tasks start from a published environment tied to GitHub repositories. Each task has its own workspace. Setup can install dependencies before the agent phase.
- As of October 30, 2026, the Codex models page lists GPT-6.1 Terra for ChatGPT Work and the Codex CLI from version 0.159.0.
Use cases
Where this shines in production.
- Local and cloud coding tasks
- CLI reviews with a selected model
- Shared project environments for parallel cloud tasks
Limitations & trade-offs
What to watch for.
- Model availability depends on the rollout, sign-in method, client, and workspace settings. Selecting a model does not grant access the plan does not already have.
- GPT-6.1 Sol Ultrafast support is documented as coming later. Standard and Fast modes are available at launch.
- This is the Codex product, not the Agents API. The Agents API is the managed harness for your own application.
Models referenced
Declared model dependencies or integrations.
GPT-6.1 Terra, GPT-6.1 Sol, GPT-6 Astra, GPT-6 Luna
Related prompts
Hand-picked or latest prompt templates.
Prompt
Prompt-injection resistant assistant
System prompt that treats retrieved docs, tickets, and web pages as untrusted data. Use in multi-tenant RAG and browsing agents. This is a defensive template, not an attack or bypass playbook.
Prompt
Computer-use agent with approval gates
System prompt for browser/desktop agents (Grok Bot, computer-use tools). Allows research and drafts, but pauses before send, purchase, delete, or production changes. Not a jailbreak or bypass guide.
Prompt
Transcript to summary and action items
Turn a speech transcript into decisions, owners, and follow-ups. Use after Gemini 3.5 Transcribe, Whisper, or MAI-Transcribe. Do not use on raw audio; transcribe first.
Prompt
Product image generation prompt
A reusable image prompt for catalog and hero shots: subject, lighting, lens, negative constraints, and text-in-image rules. Works with Firefly, Midjourney, MAI-Image, Qwen-Image, and SDXL. Not for photorealistic people or brand-logo cloning.
Prompt
LLM output evaluation rubric
Judge a model answer against a task, retrieved evidence, and pass/fail thresholds. Use for golden-set grading and regression evals. Not a substitute for human review on high-risk decisions.
Prompt
Strict JSON extraction prompt
Extract structured fields into JSON that matches a schema. Pair with provider structured-output / JSON-schema modes when available. Do not use when the output is meant to be prose.
Looking for a tighter match? Search the prompt library.
OpenAI Codex FAQ
What is OpenAI Codex?
Codex is OpenAI's coding agent for the ChatGPT desktop app, ChatGPT Work on the web, the Codex CLI, and the IDE extension. The Codex models page recommends Astra for the hardest end-to-end work, GPT-6.1 Sol for repeated long-running work, GPT-6.1 Terra for everyday writing, visual understanding, and computer use, and Luna for focused high-volume tasks. Those recommendations apply to ChatGPT Work and Codex. GPT-6.1 Terra requires Codex CLI 0.159.0 or later. Run codex -m gpt-6.1-terra. The Terra card lists Work and Codex in the desktop app, ChatGPT Work on the web, the Codex CLI, ChatGPT Credits, and API access. Codex Cloud runs tasks in a published environment with its own workspace.
When should teams use OpenAI Codex?
Local and cloud coding tasks
What should teams watch out for with OpenAI Codex?
Model availability depends on the rollout, sign-in method, client, and workspace settings. Selecting a model does not grant access the plan does not already have.
Related
Comparisons, platforms, and models teams often view next.