Codex vs OpenHands: How They Actually Differ
Pick Codex if you want a managed coding agent on OpenAI's GPT models, with work continuing across CLI, IDE, and GitHub, and a free tier included with ChatGPT. Pick OpenHands if you want an MIT-licensed, self-hostable agent you can point at any major LLM with your own key.
Compiled 2026-08-31 from our catalogue entries for Codex and OpenHands.
The short answer
- Pick Codex if
- you want a managed agent that runs the same workflow across terminal, IDE, desktop, and cloud, reviews GitHub pull requests, and is included with ChatGPT plans down to a free tier.
- Pick OpenHands if
- you want an open-source, Docker-hosted agent that accepts any major LLM via your own API key, and you are comfortable running the infrastructure yourself.
- Cheapest way in
- Codex has a published Free plan at $0/month; OpenHands' starting price is not recorded and no verified plans are on file for it.
- What we could not compare
- OpenHands pricing, hosted-tier feature limits versus Codex's cloud tier, whether Codex can be self-hosted (not mentioned in our file), and any OpenHands GitHub PR-review capability (not mentioned).
Both Codex and OpenHands sit in the AI Coding Agents category and are sold as freemium products with command-line support on Windows, macOS, and Linux. Beyond that overlap, they take opposite bets on how a coding agent should be delivered.
Codex is a managed, closed tool from OpenAI that runs on GPT models and is available across the terminal, an IDE extension, a desktop app, and a cloud environment, with a GitHub pull-request review built in. OpenHands is an MIT-licensed platform you self-host in Docker; it is model-agnostic by design, since you supply your own API key for any major LLM, and it exposes a sandboxed browser, file editor, and terminal to the agent.
Side by side
Every cell is a field on our catalogue entry for that tool, copied as recorded. Rows where the two differ come first; blank fields say so rather than being filled in.
| Codex | OpenHands | |
|---|---|---|
| Starts at | Free — $0 /month | Not recorded |
| Runs on | web, mac, win, linux, cli, vscode-ext | web, mac, win, linux, cli |
| Public source code | No public repo listed | Yes — public repo |
| Category | AI Coding Agents | AI Coding Agents |
| Pricing model | Freemium | Freemium |
| Public docs | Not listed | Not listed |
- Public source code: From the repository link on each catalogue entry, not a licence audit.
Where Codex is the better fit
- One agent that runs across CLI, IDE, desktop, and cloud, so work moves between surfaces.
- Cloud code review on GitHub pull requests that comments inline and proposes fixes.
- Reads the shared AGENTS.md standard, supports MCP tools, and is included with ChatGPT plans from the free tier up through Pro.
Where OpenHands is the better fit
- MIT-licensed and self-hostable in Docker, with a public repo under All-Hands-AI/OpenHands.
- Bring-your-own-key for any major LLM provider, so the agent is not tied to one vendor.
- Sandboxed runtime that exposes a browser, file editor, and terminal to the agent.
Which one should you pick?
It runs in terminal, IDE, and desktop, includes a $0/month tier, and is bundled with ChatGPT plans up through Pro.
It is MIT-licensed and Docker-hosted with BYO key support for any major LLM, which keeps vendor lock-in out of the stack.
A cloud code review on pull requests that proposes fixes is listed as a strength; OpenHands' file does not mention GitHub PR review.
Self-hostable in Docker with any major LLM via your own key, trading managed convenience for openness.
Frequently asked questions
- Can Codex use non-OpenAI models like Claude or Gemini?
- No. Codex is listed as locked to OpenAI's GPT models, with no Claude or Gemini support out of the box.
- Can I self-host Codex?
- Our file does not mention a self-hosting option for Codex. OpenHands, by contrast, is explicitly self-hostable in Docker.
- Does OpenHands cost anything?
- Pricing is described as freemium, but the starting price is not recorded and no verified plans are on file.
- Does either tool review GitHub pull requests?
- Codex is listed as having cloud code review on pull requests that comments inline and proposes fixes. OpenHands' file does not mention GitHub PR review.
- Do both work in the terminal?
- Yes. Both list cli among their supported platforms.
What we compared
This page states nothing that is not in the two catalogue entries below. The exact text used to write it is reproduced here so you can check our reading, and both tools link through to their own pages and to the vendor.
- /tools/codex — checked 2026-07-02 · official site
- /tools/openhands — checked 2026-05-25 · official site
Show the source text
## Codex (/tools/codex) Tagline: OpenAI's coding agent — works in the terminal, IDE, and cloud, reviews GitHub pull requests, and runs on GPT-5 models. Category: AI Coding Agents Pricing model: freemium Starts at: Free — $0 /month Platforms: web, mac, win, linux, cli, vscode-ext Public repo: none listed Docs: none listed Verified plans (read 2026-08-21): Free $0 /month | Go $8 /month | Plus $20 /month | Pro 5x $100 /month | Pro 20x $200 /month | Business / Enterprise Not published | API Key Pay only for the tokens Codex uses, based on API pricing Strengths on file: Runs one agent across CLI, IDE, desktop, and cloud — work moves between surfaces.; Strong GitHub integration: a cloud code review on a pull request that catches real bugs and proposes fixes.; Reads the shared AGENTS.md standard and supports MCP tools.; Included with ChatGPT plans, from the free tier up through Pro. Limits on file: Locked to OpenAI's GPT models — no Claude or Gemini out of the box.; Fewer built-in customization primitives (hooks, sub-agents) than some rivals. Codex is OpenAI's agentic coding tool. It runs the same agent across a terminal CLI, an IDE extension, a desktop app, and a cloud environment, so work started in one surface can continue in another. It reads the shared AGENTS.md instructions file, speaks the Model Context Protocol (MCP), and can spin up a cloud code review on a GitHub pull request that comments inline and proposes fixes. Codex runs on OpenAI's GPT models — currently GPT-5.5 as the recommended frontier model — and is included with ChatGPT plans, from the free tier up through Pro. ## OpenHands (/tools/openhands) Tagline: Open-source autonomous coding agent platform, formerly OpenDevin, by All Hands AI. Category: AI Coding Agents Pricing model: freemium Starts at: Not recorded Platforms: web, mac, win, linux, cli Public repo: https://github.com/All-Hands-AI/OpenHands Docs: none listed Verified plans: none on file Strengths on file: MIT-licensed and fully self-hostable in Docker.; Bring-your-own-key for any major LLM provider.; Sandboxed runtime with browser, file editor, and terminal exposed to the agent. Limits on file: Self-hosting requires Docker familiarity and a powerful machine.; Hosted SaaS tier still feature-limited compared to commercial agents like Devin. OpenHands is an open-source autonomous coding-agent platform from All Hands AI — formerly OpenDevin. Its edge is an open, self-hostable take on the autonomous agent: MIT-licensed and self-hostable in Docker, bring-your-own-key for any major LLM, and a sandboxed runtime that exposes a browser, file editor, and terminal to the agent. It's for teams who want Devin-style autonomy without the lock-in and are willing to run it themselves. Caveats: self-hosting needs Docker familiarity and a capable machine, and the hosted SaaS tier is still feature-limited compared with commercial agents like Devin. It's the open, self-hostable counterpart to the managed, closed Devin.
Change log
- 2026-08-31 — First published, compiled from the catalogue entries for Codex and OpenHands.