OpenAI Codex pricing and plans

Jump to plans · which plan fits · what uses your allowance · usage by model · changes.

Every plan that includes Codex

Plan	Price	What you get with Codex
Free	$0	Quick coding tasks; GPT-6 Luna at Standard speed in the desktop app, subject to rollout
Go	$8/month	Lightweight coding tasks; GPT-6 Luna at Standard speed in the desktop app, subject to rollout
Plus	$20/month	Codex on the web, CLI, IDE extension, and iOS; cloud code review and Slack; GPT-6.1 Sol and Terra, GPT-6 Sol and Luna, plus limited GPT-6 Astra; buy credits when you hit the limit
Pro	$100, $200, or $500/month	No five-hour limit (weekly limits may apply). Pro 200 includes more usage than Pro 100. Pro 500 has OpenAI’s highest allowance and is the only Pro plan with GPT-6 Astra and GPT-6.1 Sol Ultrafast. On October 29, OpenAI doubled the Work and Codex allowance on Pro 100 and Pro 500 and canceled the planned Pro 200 cut, so every Pro 200 subscription has the higher allowance
Business	Standard seat $20/user/month billed annually ($25 monthly); Premium seat $100 annually ($125 monthly). 2+ seats	Admin controls, SAML SSO, larger cloud machines, no training on business data by default. Premium seats include higher Work and Codex usage than Standard seats, raised on October 29, and no five-hour limit
Enterprise and Edu	Contact sales	Priority processing, SCIM, EKM, RBAC, audit logs, data residency
API key	Pay per token at API rates	CLI, SDK, and IDE extension only; no cloud features

Which plan fits

• Trying Codex: Free or Go.
• A few focused coding sessions a week: Plus at $20, the same price as Claude Pro and Windsurf Pro. When you hit the limit, buying credits can be cheaper than upgrading.
• Running out of Plus regularly: Pro 100 at $100, with more usage than Plus and no five-hour limit. Move to Pro 200 only if Pro 100 keeps interrupting work.
• Needing Ultrafast output: Pro 500, the only individual plan with GPT-6 Astra and GPT-6.1 Sol Ultrafast. Ultrafast uses your allowance 8x faster than Standard.
• A team: Business at $20 per user billed annually, close to GitHub Copilot Business at $19 per seat. Heavy users can take a Premium seat at $100 per user billed annually.
• CI jobs and automation: an API key, billed per token with no cloud features.

What uses your allowance faster

OpenAI says tasks that look similar can use very different amounts. Model choice, context size, reasoning, tool use, retrieval, and caching all count, so prompt length alone is not a reliable guide.

• Cloud chats can use more of your allowance than local messages. Local and cloud work share one allowance.
• Fast mode uses included limits at 2.5x the Standard rate, and purchased credits at 2x.
• GPT-6 Astra and GPT-6.1 Sol Ultrafast, on Pro 500 and eligible Enterprise and Edu plans, uses included limits at 8x and purchased credits at 6x.
• Image generation uses included limits about 3–5x faster than similar turns without images.
• Voice in the desktop app draws $0.05 per minute from your Codex usage budget.
• ChatGPT Work and some agentic features, including ChatGPT for Excel on Plus and Pro, share the same allowance.

When you hit the limit: work already in progress can finish its current turn. After that, Plus and Pro users can buy credits, Business and Enterprise workspaces with flexible pricing can buy workspace credits, and anyone can run extra local chats with an API key at API rates. To stretch your allowance, use a smaller model for routine tasks, keep prompts and AGENTS.md files short, and turn off MCP servers you don’t need. For how credits, banked resets, and paid resets work, see Codex usage limits explained: plan allowance, credits, and resets.

Usage estimates by model

OpenAI publishes ranges of local messages per five-hour window for Plus and Standard Business, not fixed caps. Pro plans currently have no five-hour limit, so these ranges don’t apply to Pro. The top model, GPT-6 Astra, gets the smallest range.

Model	Local messages per 5 hours (Plus and Standard Business)
GPT-6 Astra (most capable)	5–45
GPT-6.1 Sol	15–160
GPT-6.1 Terra	35–285
GPT-6 Sol	15–150
GPT-6 Luna	350–3,000

Beyond the included allowance, usage is charged in credits per million tokens. The gap between models is large: GPT-6 Astra costs 100 times as much per token as GPT-6 Luna.

Model	Input	Cached input	Output
GPT-6 Astra	250	25	1,250
GPT-6.1 Sol	50	2.5	250
GPT-6.1 Terra	25	1.25	125
GPT-6 Sol	50	5	250
GPT-6 Luna	2.5	0.25	12.5
GPT-5.6 Sol	100	10	500
GPT-5.6 Terra	50	5	300
GPT-5.6 Luna	5	0.5	30

Recent and upcoming changes

Ultrafast scope rechecked October 26, 2026: both GPT-6 Astra and GPT-6.1 Sol use 8× included usage and 6× purchased credits, relative to Standard for the same model. Official speed-mode rules.

Recent changes

• October 29, 2026: OpenAI doubled the Work and Codex allowance on Pro 100 and Pro 500 and canceled the Pro 200 reduction it had announced on September 29. Pro 200 subscribers on the lower allowance were raised to the higher one, new subscriptions get it too, and prices didn’t change. Business Premium usage also increased; OpenAI gave no multiple. Official release note; checked November 5, 2026. ChatGPT Pro Tiers for Codex: Usage, Access, and Limits.
• October 26, 2026: OpenAI made GPT-6.1 Terra available in ChatGPT Work and Codex on Plus, Pro, Business, Enterprise and Edu; it isn’t available in Chat. Its Codex credit rates are half of GPT-6.1 Sol’s (25 input, 1.25 cached input and 125 output per million tokens), and Plus gets 35–285 local messages per five hours. OpenAI says overall Work and Codex plan limits are unchanged. GPT-6.1 Terra Pricing Explained: API Costs, Codex Usage, and Who It Fits.
• October 14, 2026: OpenAI’s scheduled GPT-5.5 retirement covers ChatGPT, ChatGPT Work and Codex on all plans; the OpenAI API is unaffected. For ChatGPT sign-in, select a replacement available to your plan and client, and update saved model settings, workspace defaults, agents, scheduled tasks and scripts that still select gpt-5.5. The notice does not give a cutoff time. Official retirement notice; checked October 14, 2026.
• September 29, 2026: OpenAI launched dots, always-on agents powered by GPT-6 Astra, with the first dot included in eligible Pro and Business Premium plans. Talking to a dot doesn’t count toward ChatGPT limits, but Codex tasks it starts or manages count toward your Codex limits as usual. OpenAI Dots: pricing, included plans, and what counts toward usage.
• September 29, 2026: OpenAI released GPT-6.1 Sol. Its API base rates match GPT-6 Sol ($2 input, $10 output per million tokens), and cached input falls from $0.20 to $0.10. In Codex, its cached input costs 2.5 credits per million, half GPT-6 Sol’s rate, and Plus gets 15–160 local messages per five hours. GPT-6.1 Sol pricing: what changed and what Codex users should know.
• September 29, 2026: OpenAI added Pro 500 at $500 a month, with GPT-6 Astra Ultrafast, and reopened Pro 200 to new subscribers at a lower allowance than before. OpenAI said eligible existing Pro 200 subscribers would keep their previous allowance through October 29; that reduction was later canceled (see October 29). Pro plans currently have no five-hour limit; weekly limits may still apply. ChatGPT Pro $100 vs $200 vs $500 for Codex: what OpenAI has confirmed.
• September 10, 2026: OpenAI paused new sign-ups and upgrades to Pro 200. The pause ended with the September 29 reopening.

Upcoming dates

• Through at least November 21, 2026: GPT-5.6 Sol keeps its promotional pricing.

Related guides

• Codex usage limits explained: plan allowance, credits, and resets
• OpenAI Dots: pricing, included plans, and usage
• ChatGPT Pro $100 vs $200 vs $500 for Codex
• GPT-6.1 Sol pricing: what changed and what Codex users should know
• GPT-6.1 Terra pricing: API costs, Codex usage, and who it fits
• AI coding tools pricing compared
• Cursor Pro vs GitHub Copilot Pro
• Best GitHub Copilot alternatives
• How Toolvel checks prices