All vikis
007 First Light
AI
Aion 2
Ananta
ARC Raiders
ArcheAge Chronicles
Call of Duty: Modern Warfare 4
Chrono Odyssey
Civilization 3
Civilization 4
Claude
Crimson Desert
Deadlock
Diablo 4
Diablo 5
Enshrouded
Fable
Final Fantasy Resonance
Football Manager 27
Forza Horizon 6
God of War
Grand Theft Auto 6
Heroes of Might and Magic 3Codex Pricing and Limits
Codex comes with every ChatGPT plan as of Nov 6, 2026: Free ($0) and Go ($8) get GPT-6 Luna in the desktop app, Plus ($20) gets about 15 to 160 GPT-6.1 Sol messages per five hours plus Codex Cloud, and Pro ($100, $200 or $500) has no five-hour limit. Credits extend any plan, and an API key bills by the token instead.
Codex costs nothing extra on top of a ChatGPT plan, and as of Nov 6, 2026 every plan includes some: Free ($0) and Go ($8 a month) get GPT-6 Luna in the desktop app, Plus ($20 a month) gets roughly 15 to 160 local messages every five hours with GPT-6.1 Sol plus Codex Cloud and GitHub code review, and Pro ($100, $200 or $500 a month) has no five-hour limit. When the included usage runs out, Plus and Pro can buy credits, and anyone can keep working with an OpenAI API key at API prices. Codex usage is shared with ChatGPT Work, so a long Work task and a long Codex task draw on the same allowance (pricing).
Plan prices in general are on ChatGPT Plans and Pricing and the credit system on ChatGPT Credits; this page is only about what Codex gets.
Quick facts
| Free | $0: GPT-6 Luna at Standard speed in the desktop app, subject to rollout |
| Go | $8 a month: the same, "for lightweight coding tasks" |
| Plus | $20 a month: Codex everywhere, Cloud, code review, GPT-6.1 Sol and GPT-6 Luna, limited GPT-6 Astra |
| Pro | $100, $200 or $500 a month; no five-hour limit; Ultrafast only on Pro $500 |
| Business | $20 per user a month billed annually, $25 monthly |
| Enterprise and Edu | Contact sales |
| Shared with | ChatGPT Work, and ChatGPT for Excel on Plus and Pro |
| Usage dashboard | chatgpt.com/codex/settings/usage |
Prices in US dollars, as of Nov 6, 2026.
What does each plan include?
| Plan | Price | Codex | Cloud and integrations |
|---|---|---|---|
| Free | $0 | GPT-6 Luna in the desktop app, subject to rollout; "Explore Codex capabilities on quick coding tasks" | No |
| Go | $8 a month | GPT-6 Luna in the desktop app, subject to rollout; "lightweight coding tasks" | No |
| Plus | $20 a month | Web, CLI, IDE extension, iOS; GPT-6.1 Sol, GPT-6 Luna; limited GPT-6 Astra | Cloud, automatic code review, Slack |
| Pro 100 / 200 / 500 | $100 / $200 / $500 a month | No five-hour limit; full allowance usable for Astra on Pro $100 and $200; Ultrafast on Pro $500 | Yes |
| Business | $20 per user a month (annual), $25 (monthly) | Standard seats get Plus-level limits; Premium seats use their full allowance for Astra | Yes, with larger cloud machines |
| Enterprise and Edu | Contact sales | Plus-level limits per seat, or no fixed limits with flexible pricing | Yes |
Source: Pricing, OpenAI docs; ChatGPT Work and Codex and Using Codex with your ChatGPT plan, OpenAI Help Center; read Nov 6, 2026. Free and Go do not include Codex Cloud. Image generation in Codex is not available on Free (pricing).
How many Codex messages do I get?
OpenAI publishes ranges, not fixed numbers, for local messages (in the app, CLI or IDE) every five hours on Plus and on Standard Business seats. How many you actually get depends on the size of the task and the code; "weekly limits may also apply", and cloud tasks may use more (pricing).
| Model | Local messages per five hours, Plus and Standard Business |
|---|---|
| GPT-6 Astra | 5 to 45 |
| GPT-6.1 Sol | 15 to 160 |
| GPT-6.1 Terra | 35 to 285 |
| GPT-6 Sol | 15 to 150 |
| GPT-6 Luna | 350 to 3,000 |
Source: Pricing, OpenAI docs, read Nov 6, 2026.
OpenAI gives no numbers for Free or Go. Pro plans "currently have no five-hour limit"; their allowances are described only in relative terms: on Oct 29, 2026 Pro $100 and Pro $500 got "twice their previous Work and Codex allowance", and on Pro $200 only the lower allowance doubled, while "Plus usage is unchanged" (About ChatGPT Pro tiers). Enterprise and Edu workspaces with flexible pricing have no fixed rate limits; usage scales with credits (pricing).
To make a limit last, OpenAI suggests a smaller AGENTS.md, fewer MCP servers, and a lighter model such as GPT-6 Luna for simple work (pricing). Type /status in the CLI, or open the usage dashboard, to see what is left.
Fast and Ultrafast
Speed costs usage. Fast mode spends included usage 2.5 times as fast and purchased credits at twice the rate. Ultrafast, offered for GPT-6 Astra and GPT-6.1 Sol, uses 8 times the included usage and 6 times the credits (pricing). Among personal plans only Pro $500 has Ultrafast; it is also on eligible Enterprise and Edu plans, and not on Plus, Pro $100, Pro $200 or Business, even with credits (ChatGPT Work and Codex). OpenAI's Sep 29, 2026 video announced Ultrafast for GPT-6 Astra in the API, ChatGPT Work and Codex, and GPT-6.1 Sol Ultrafast launched on Oct 8, 2026. Turn Fast on or off in the CLI with /fast.
Credits
When the included usage runs out, credits pay for more. Personal plans use included usage first, then credits; the same credits cover Codex, ChatGPT Work and ChatGPT for Word, Excel and PowerPoint. Credits last 12 months from purchase, can reload automatically, and are not API credits. Buying credits is open to Plus and Pro, and on Free and Go only to "a limited group" (Using credits). OpenAI does not publish a single dollar price per credit; amounts and prices "can vary by account, region, and plan".
A task costs credits by the tokens it uses. Rates per million tokens (input / cached input / output) at Standard speed, as of Nov 6, 2026:
| Model | Credits per 1M tokens |
|---|---|
| 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 |
Source: Pricing, OpenAI docs, read Nov 6, 2026. OpenAI says a typical GPT-5.6 Sol task "may use 2-15 credits", and GPT-6.1 Sol drawn from a subscription has no long-context surcharge. Business, Edu and Enterprise workspaces on credit billing use a separate rate card (pricing).
Other things that use Codex usage
- Voice in the desktop app costs $0.05 a minute of Codex usage, or 1.25 credits a minute on credit-billed business plans; it does not work with an API key.
- Image generation in Codex uses included limits "3-5x faster on average" (pricing).
- Code review uses Codex usage only when it runs through GitHub (a
@codex reviewtag or automatic reviews) (pricing); see Codex Code Review. - Auto-review safety checks are free and do not count toward limits (OpenAI Help Center), though OpenAI's docs also say auto-review "uses extra model calls" (Agent approvals and security).
- Composer predictions on Pro do not count toward limits (Composer predictions).
What happens when I hit the limit?
Work already in a turn may finish, "subject to fair use limits". After that, Plus and Pro can buy credits, workspaces with flexible pricing buy workspace credits, and anyone can switch the CLI, IDE extension or SDK to an API key and pay API rates (pricing). OpenAI support cannot reset limits (About ChatGPT Pro tiers). Since Jun 11, 2026 eligible Plus and Pro users can bank rate-limit resets, starting with one free reset (release notes).
Using Codex with an API key
With an API key, Codex runs in the CLI, the SDK and the IDE extension and bills at API prices, with the models your key can reach. There are no cloud features (no GitHub code review, no Slack) (pricing). This suits CI jobs and teams already paying for the API; see Codex SDK.
Students
Since Mar 20, 2026, verified university students in the United States and Canada can claim "$100 in ChatGPT credits, or 2,500 credits" for Codex beyond their plan's limits, at chatgpt.com/codex/students with SheerID verification; the credits expire 12 months after the grant and are not API credits (release notes). Open-source maintainers can apply to Codex for OSS, which OpenAI describes as free ChatGPT Pro and Plus accounts, code review and Codex Security (OpenAI, Mar 6, 2026). Student plans in general are on ChatGPT for Education.
What testers found
In an Aug 27, 2026 video comparing Codex with Claude Code, Tim Ruscica of Tech With Tim pointed to what the entry price buys: ChatGPT Plus at $20 included OpenAI's flagship model in Codex, while Claude's $20 plan defaulted to its middle model, with the flagship from the $100 plan up.
On limits, he said that at the time of filming Codex was "clearly much more generous" than Claude Code. The Plus message counts he quoted came from community trackers he called hard to verify, and in his own use similar tasks drew about 3 to 4 times less of his Codex allowance. Those were the GPT-5.6-era limits; OpenAI's current ranges are in the table above.
Frequently asked questions
How much does Codex cost? Codex has no separate price: it comes with ChatGPT plans, from Free ($0) and Go ($8 a month) with GPT-6 Luna, to Plus ($20) with Codex Cloud, and Pro ($100, $200 or $500) with no five-hour limit, as of Nov 6, 2026.
What are the Codex usage limits on Plus? On ChatGPT Plus, Codex allows roughly 15 to 160 local GPT-6.1 Sol messages, 5 to 45 GPT-6 Astra messages or 350 to 3,000 GPT-6 Luna messages every five hours, and weekly limits may also apply. These are OpenAI's estimates, not fixed numbers.
Does ChatGPT Pro have Codex limits? Pro has no five-hour Codex limit as of Nov 6, 2026, but it is not unlimited: allowances apply, and Pro $500 has the largest. OpenAI publishes no message numbers for Pro.
Can I buy more Codex usage? Yes. ChatGPT Plus and Pro users can buy credits that last 12 months and cover Codex, ChatGPT Work and ChatGPT for Office apps; anyone can also run Codex with an API key at API prices.
Is Codex Cloud included with ChatGPT Go? No. Codex Cloud needs ChatGPT Plus or higher; Free and Go get Codex only in the desktop app with GPT-6 Luna, subject to rollout.
Do Codex and ChatGPT Work share limits? Yes. ChatGPT Work and Codex share one usage allowance and the same credits, so time spent in one leaves less for the other.
Sources
- Pricing, OpenAI docs, read Nov 6, 2026
- Using Codex with your ChatGPT plan, OpenAI Help Center, read Nov 6, 2026
- ChatGPT Work and Codex, OpenAI Help Center, read Nov 6, 2026
- About ChatGPT Pro tiers, OpenAI Help Center, read Nov 6, 2026
- Using credits for flexible usage in ChatGPT personal plans, OpenAI Help Center, read Nov 6, 2026
- ChatGPT release notes, OpenAI Help Center, read Nov 6, 2026
- Changelog, OpenAI docs, read Nov 6, 2026
- Agent approvals and security, OpenAI docs, read Nov 6, 2026
- Composer predictions, OpenAI docs, read Nov 6, 2026
- Codex Security: now in research preview, OpenAI, Mar 6, 2026
- Build faster with Ultrafast, OpenAI, Sep 29, 2026 (video)
- Codex vs Claude - an Honest Comparison, Tech With Tim, Aug 27, 2026 (video)
Moments
Build faster with Ultrafast
OpenAI
Codex vs Claude - an Honest Comparison
Tech With Tim
Related
-
-
-
-
-
Codex ModelsAs of Nov 6, 2026 Codex recommends four models: GPT-6 Astra (most capable), GPT-6.1 Sol (the default, near Astra at lower cost), GPT-6.1 Terra (fast, for writing and computer use) and GPT-6 Luna (most efficient). They replaced a line of coding-tuned models that ran from codex-1 in 2025 through GPT-5-Codex, GPT-5.1-Codex-Max and GPT-5.3-Codex.
- Codex SDKThe Codex SDK lets programs run OpenAI's Codex agent: npm install @openai/codex-sdk for TypeScript (Node.js 18 or later) or pip install openai-codex for Python (3.10 or later). For CI there is also codex exec, the non-interactive CLI mode, and the GitHub Action openai/codex-action@v1; the App Server is the protocol for building full Codex clients.
- Codex SecurityCodex Security is OpenAI's application security agent: it builds a threat model of a repository, looks for vulnerabilities, tries to confirm them in a sandbox and proposes patches. It began as Aardvark (announced Oct 30, 2025), became Codex Security in research preview on Mar 6, 2026, and now runs as a desktop plugin, a cloud plugin, a CLI and SDK, and pull request Security Review, alongside Daybreak for approved defenders.
-
ChatGPT Atlas vs Other AI BrowsersChatGPT Atlas, OpenAI's AI browser, stopped working on Aug 9, 2026; ChatGPT's browsing now lives in the desktop app's built-in browser and the ChatGPT browser extension, and the AI browsers still sold are Perplexity's Comet (free), Gemini in Chrome, Edge with Copilot, Dia and Claude in Chrome.
See also
-
-
-
ChatGPT Plans and PricingChatGPT costs nothing on Free, $8 a month on Go, $20 on Plus and $100, $200 or $500 on Pro in the US as of Nov 6, 2026, with Business at $20 to $25 per user a month and Enterprise and Edu priced by OpenAI's sales team.
-
ChatGPT CreditsChatGPT credits buy more Codex, ChatGPT Work and Office extension use after a plan's included usage runs out; as of Nov 6, 2026 they are charged per million tokens by model, last 12 months, and are separate from the API credits that come with Pro and Business.
-
CodexCodex is OpenAI's coding agent: it reads a codebase, edits files, runs commands and tests, reviews pull requests and works in the cloud. As of Nov 6, 2026 it lives inside the ChatGPT desktop app, in the Codex CLI, in IDE extensions, in Codex Cloud and on mobile, and every ChatGPT plan includes some of it, Free and Go too.
-
AGENTS.mdAGENTS.md is a plain Markdown file of instructions for AI coding agents, a "README for agents", that Codex reads before doing any work. Codex loads a global ~/.codex/AGENTS.md and then every AGENTS.md from the repository root down to the working folder, closest last, up to 32 KiB; /init writes a first one. The open format is now stewarded by the Agentic AI Foundation.