Claude Code vs OpenAI Codex (2026): Plans, Limits and Workflow
Prices verified against official sites on 2026-10-21 · Rankings are independent of sponsorships
Claude Code and OpenAI Codex are the two coding agents most developers compare first. Both come bundled with a chat subscription you may already pay for, both run in the terminal and the IDE, and both can bill API usage instead. The real differences are which plan unlocks what, how the usage limits behave, and how much of the work happens in the cloud. Everything below comes from claude.com/pricing and OpenAI's Codex pricing page as we read them on October 21, 2026.
Quick verdict
At a glance
| Tool | Tier | Pricing | Free tier |
|---|---|---|---|
| Claude Code | S | Included with Claude Pro ($20/mo, or $17/mo billed annually), Max ($100/mo, with $100/mo in Claude Platform API credits) and Ultra ($200/mo, formerly Max 20x, with $200/mo in API credits and no 5-hour limit for verified accounts); Team seats $25/mo standard, $125/mo premium ($20/$100 billed annually); or pay-as-you-go via the API. | Not included in the Free Claude plan |
| OpenAI Codex | A | Included in ChatGPT Free, Go ($8/mo), Plus ($20/mo), Pro ($100, $200 or $500/mo), Business, Edu and Enterprise; extra usage via ChatGPT credits. | Yes, ChatGPT Free includes Codex for quick tasks |
| GitHub Copilot | A | Free $0; Pro $10/mo; Pro+ $39/mo; Max $100/mo. Chat and agents use GitHub AI Credits (1 credit = $0.01): Pro includes $15/mo of credits, Pro+ $70/mo, Max $200/mo; completions unlimited on paid plans. | Yes, Copilot Free: 2,000 completions/month plus limited chat and agent usage (includes CLI and agent mode) |
The short answer
Pick Codex if you want to start free or cheap. ChatGPT Free and Go ($8/mo) include Codex for quick tasks, and Plus ($20/mo) adds Codex on the web, in the CLI, in the IDE extension and on iOS, plus automatic code review on GitHub and a Slack integration.
Pick Claude Code if you mainly delegate long, multi-file tasks from the terminal and want the subagent, hooks and CLAUDE.md setup it is known for. It starts at Claude Pro ($20/mo, or $17/mo billed annually) and is not included in Claude Free.
Pricing side by side (verified Oct 21, 2026)
Claude Code: not on Free; Pro $20/mo ($17/mo billed annually, $200 up front); Max $100/mo (5x Pro usage) and Ultra $200/mo (Anthropic renamed Max 5x to Max and Max 20x to Ultra on Oct 30, 2026, at the same prices); Team standard seats $25/seat/mo ($20 billed annually) and premium seats $125/seat/mo ($100 billed annually), for teams of 2 to 150; Enterprise $20/seat/mo billed annually plus usage at API rates (20-seat minimum).
OpenAI Codex: Free $0 and Go $8/mo (GPT-6 Luna at Standard speed in the desktop app, subject to rollout); Plus $20/mo; Pro at $100, $200 or $500/mo; Business $25/user/mo billed monthly (2+ users); Edu and Enterprise via sales. You can also run Codex with an API key and pay API prices.
How the usage limits work
Claude: every plan's limits reset on a rolling five-hour window, and paid plans add weekly limits. Claude Code shares one pool with your chats, so a heavy coding day eats into chat usage too. Pro gives at least 5x the per-session usage of Free, Max gives 5x Pro every 5 hours and every week, and Ultra (formerly Max 20x) gives 10x Pro weekly usage with no 5-hour limit for verified accounts (as of Oct 30, 2026). When you hit a limit you can wait, upgrade, or turn on usage credits billed at standard API rates.
Codex: usage depends on the model and the task, not a fixed message count. OpenAI's estimate for Plus is 15 to 160 local messages per five hours on GPT-6.1 Sol and 350 to 3,000 on GPT-6 Luna, and cloud tasks can use more. Pro plans currently have no five-hour limit. Plus and Pro users who run out can buy ChatGPT credits instead of upgrading.
Workflow: terminal agent vs cloud agent
Claude Code is agent-first in the terminal, with IDE and web surfaces. It does best when you hand it a whole task: plan, edit many files, run tests and iterate, steered by CLAUDE.md memory, subagents and hooks.
Codex spans the ChatGPT desktop app, the CLI, the IDE extension, the web and iOS. On Plus and above it also runs cloud tasks, takes work from GitHub issues and PRs when you tag @codex, reviews pull requests automatically, and connects to Slack and Linear. With API-key sign-in you keep the CLI, SDK and IDE extension but lose the cloud features.
Teams and API-key use
For a five-developer team, Claude Team standard seats work out to $125/mo billed monthly (5 x $25), or $625/mo if everyone needs premium seats. Codex on ChatGPT Business is $125/mo billed monthly for the same five people (5 x $25). Both add SSO and central admin; check each vendor's page for the exact feature split.
Both tools also run on pay-as-you-go API keys, which suits CI and shared automation. Claude Code uses a Claude Console account; Codex with an API key works in the CLI, SDK and IDE extension, with model availability following your API key.
Or use both from GitHub Copilot
GitHub Copilot (Free, Pro $10/mo, Pro+ $39/mo, Max $100/mo) now lets you delegate tasks to third-party coding agents, including Claude by Anthropic and OpenAI Codex, from its cloud agent (in preview). If your team already lives in GitHub, that's a low-friction way to try both before committing to one subscription.
Whichever agent you pick, the free CLAUDE.md and AGENTS.md templates in the ToolTier Library give it project context from day one. Codex reads AGENTS.md, and Claude Code reads CLAUDE.md.
Make these tools better
- Claude Code Agent Kit ($19): Installable Claude Code plugin: 12 subagents, 8 skills and 3 tested safety hooks.
- Vibe Debug Pack ($9): 42 symptom prompts, a decision tree and git rescue to break AI fix loops.
Prefer free first? Browse the ToolTier Library for copy-paste Cursor rules, CLAUDE.md, AGENTS.md and debug prompts.
Prices change often. Always confirm on the vendor's site before buying. See the full tier lists.
