- What is the best alternative to Codex CLI?
- Warp is the one we can answer for in full. Keep the CLI agent you already use and change only the terminal: Warp runs Claude Code, Codex and Gemini CLI beside its own agent. Warp is $0 on Free, $20 per month on Build, $200 per month on Max and $50 per user per month on Business; annual billing takes Build to $18 and Max to $180. Every part of that answer is quoted from a page we read; the links are on this page beside each claim.
- Which Codex CLI alternatives are still actively developed?
- Warp (newest commit 8 Oct 2026), Claude Code (newest commit 8 Oct 2026), Goose (newest commit 7 Oct 2026), Qwen Code (newest commit 8 Oct 2026). Those dates are the newest commit on each project's default branch, read from the GitHub API on 8 Oct 2026. We read the default branch rather than the repository's push timestamp, because a push to any branch moves the second one and makes a finished project look active.
- What do Codex CLI alternatives cost?
- Warp: Warp is $0 on Free, $20 per month on Build, $200 per month on Max and $50 per user per month on Business; annual billing takes Build to $18 and Max to $180. Claude Code: Claude Pro costs $17 per month on the annual plan, $200 billed up front, or $20 per month billed monthly; Max starts at $100 per month. opencode: OpenCode Go, the optional subscription, costs $5 for the first month and $10 a month after that, and one workspace can hold only one subscriber. Crush: Charm Hyper, the provider built for Crush, gives 100 Hypercredits a month at $0 and 250 a day on a $20 monthly subscription; 1 Hypercredit is 5 cents. Prices we have not read ourselves are marked "not verified" in the table above rather than copied from our catalogue, which is out of date on 43 of our 190 rows.
- How many Codex CLI alternatives do you track?
- We name 17 on this page. 8 carry judgments and prices we read and quoted, 3 carry live repository data with no judgment yet, and 6 are catalogue entries we have not researched. We say which is which on every row rather than presenting all of them as reviewed.