Codex usage comes out of an allowance included in your ChatGPT plan, shared with ChatGPT Work, and measured over a five-hour window and a weekly window where your plan has both; Pro plans currently have no five-hour limit (Pricing, OpenAI Help Center). To see what is left from the terminal, type this in an active Codex CLI session:
/status
Applies to: OpenAI Codex with ChatGPT sign-in or an API key, per OpenAI’s pricing docs and Help Center as of October 2, 2026. Last checked: October 2, 2026.
On this page
How Codex usage is counted
OpenAI’s pricing page states three rules that decide how fast an allowance goes:
- One pool. “ChatGPT Work and Codex share usage.” Local messages and cloud chats also share your plan’s allowance. The Help Center article on Codex adds ChatGPT for Excel and Workspace Agents to the same allowance and credit pool where those features are on your plan.
- No fixed message count. “The number of messages you can send depends on the model used, size and complexity of your tasks, and whether you run them locally or in the cloud.” Model choice, context, reasoning, tool use, retrieval and caching all affect usage, so prompt length is not a reliable estimate.
- Cloud costs more. “Cloud tasks may use more of your allowance than local messages.”
Image generation, Fast mode and Voice also draw on the same limits (details below).
The five-hour window and the weekly window
OpenAI’s Help Center article on managing usage describes the two windows this way:
| Window | OpenAI’s wording |
|---|---|
| Five-hour | “The five-hour limit determines how much usage is included in each window. A new window starts when you send your first message in Work or Codex after the previous window ends.” |
| Weekly | “The weekly limit controls how much included work you can use over the weekly usage period.” |
“Where both apply, you need allowance remaining in both to continue,” and “You may reach the five-hour limit before five hours have passed, even if you still have weekly usage remaining.” Which windows apply depends on the plan:
| Plan | Five-hour window | Weekly window |
|---|---|---|
| Plus, Standard Business | Yes; OpenAI publishes estimated local messages per five-hour period | “Weekly limits may also apply” |
| Pro ($100, $200, $500) | “Pro plans currently have no five-hour limit” | “Weekly limits may also apply” |
| Enterprise/Edu with flexible pricing | “No fixed rate limits. Usage scales with credits.” | Not applicable |
| Enterprise/Edu without flexible pricing | “Same per-seat usage limits as Plus for most features” | Same as Plus |
| API key | No plan windows; billed at API rates | Not applicable |
Sources: Pricing, Managing usage in Work and Codex.
Estimated messages per five-hour window
The pricing page gives these ranges of local messages per five-hour period. OpenAI says “These estimates are not fixed message limits; check your usage dashboard for current limits and reset times.”
| Model | Plus | Standard Business |
|---|---|---|
| GPT-6 Astra | 5-45 | 5-45 |
| GPT-6.1 Sol | 15-160 | 15-160 |
| GPT-6 Sol | 15-150 | 15-150 |
| GPT-6 Luna | 350-3,000 | 350-3,000 |
The Codex pricing page on chatgpt.com lists the same four rows plus GPT-5.6 Sol (10-100), GPT-5.6 Luna (250-2,000) and GPT-5.5 (15-80) for Plus and Standard Business. GPT-5.5 retires from Codex with ChatGPT sign-in on October 14, 2026 (Models).
On GPT-6 Astra, the Codex pricing page says it is “included within the existing Codex usage allowance on the $100 and $200 Pro plans and the $100 Business plan,” while “Plus and the $20 Business plan include limited GPT-6 Astra usage, with the option to purchase credits for additional access” (Codex pricing).
What uses the allowance faster
| Setting or feature | Effect on included usage, per OpenAI |
|---|---|
| Fast mode | 2.5x included subscription usage (2x for purchased credits) |
| GPT-6 Astra Ultrafast | 8x included subscription usage (6x for purchased credits) |
| Image generation | Uses included limits about 3-5x faster on average |
| Voice in the desktop app | Uses your Codex usage budget at $0.05 per minute |
| Higher reasoning effort | “Takes longer and uses more tokens” |
Sources: Pricing, Models. OpenAI notes the Fast and Ultrafast multipliers “don’t describe speed increases.”
How to check remaining Codex usage
OpenAI documents three places:
- In the Codex CLI: run
/statusduring an active session. The pricing page says: “If you want to see your remaining limits during an active Codex CLI session, you can use/status.” - In the usage dashboard: “You can find your current limits in the usage dashboard” (Pricing, Codex pricing).
- In the ChatGPT apps: open Settings → Usage for “your remaining allowance, usage windows and reset times” (Help Center).
OpenAI suggests checking the dashboard “every week or two to understand your pace and remaining capacity” (Pricing). Members of eligible Enterprise workspaces with credit-based billing get Settings > Usage & billing in ChatGPT Desktop, with their monthly Work and Codex allowance and recent credit history (Help Center).
What happens when you hit the limit
From the pricing page:
- “If you reach your usage limits during an active turn, the agent will be able to continue working on that turn, subject to fair use limits.”
- “ChatGPT Plus and Pro users who reach their usage limit can purchase additional credits to continue working without needing to upgrade their existing plan.”
- “Business, Edu, and Enterprise plans with flexible pricing can purchase additional workspace credits to continue working.”
- “All users may also run extra local chats using an API key, with usage charged at standard API rates.”
The Help Center adds that “Switching models does not restore allowance in a shared usage pool,” and the Pro tiers article says OpenAI Support does not reset ChatGPT or Codex usage limits.
Resets
OpenAI’s Help Center describes three kinds:
| Reset type | What it does |
|---|---|
| Banked reset | Saved to your account until used or expired. A full banked reset refreshes the five-hour and weekly limits and changes your weekly reset date. Apply it from Settings → Usage in the ChatGPT desktop app. |
| Automatic or global reset | Refreshes the limits named in OpenAI’s announcement; nothing to apply, and it does not appear as a saved reset. |
| Purchased instant reset | For eligible personal Plus and Pro accounts, depending on account and billing country. Refreshes both windows at checkout and cannot be saved. Your next weekly period starts with your first request after the reset. |
Credits
“Credits are the unit used to pay for eligible usage on credit-based plans. After you reach your included limits, available credits let you continue working. Credit purchase prices and applicable discounts depend on your plan or agreement” (Pricing). The page publishes a token rate card in credits per million tokens at Standard speed:
| Model | Input | Cached input | Output |
|---|---|---|---|
| GPT-6 Astra | 250 credits | 25 credits | 1,250 credits |
| GPT-6.1 Sol | 50 credits | 2.5 credits | 250 credits |
| GPT-6 Sol | 50 credits | 5 credits | 250 credits |
| GPT-6 Luna | 2.5 credits | 0.25 credits | 12.5 credits |
OpenAI also states that “Credit prices alone don’t determine included subscription usage” and that “API token prices are separate from subscription usage; don’t use them to estimate included tasks” (Pricing). Credits pay for usage but “do not add model access” (Help Center).
Codex usage with an API key
Signing in with an API key replaces plan limits with per-token billing. The pricing page lists the API Key option as “Codex in the CLI, SDK, or IDE extension,” “No cloud-based features (GitHub code review, Slack, etc.),” “Model availability follows the API models available to your key,” and “Pay for Codex usage based on API pricing.” The Authentication page adds: “When you sign in with an API key, Codex uses standard API pricing instead of included ChatGPT plan credits.” Per-token prices are on OpenAI’s API pricing page.
printenv OPENAI_API_KEY | codex login --with-api-key
Codex plans and their usage
| Plan | Price on OpenAI’s page | Codex usage summary | Listing |
|---|---|---|---|
| Free | $0/month | GPT-6 Luna at Standard speed in the desktop app, subject to rollout | /plans/codex-free |
| Go | $8/month | GPT-6 Luna at Standard speed in the desktop app, subject to rollout | /plans/codex-go |
| Plus | $20/month | “Power a few focused coding sessions each week”; five-hour ranges above; credits | /plans/codex-plus |
| Pro | From $100/month ($100, $200 or $500) | “No five-hour limit”; weekly limits may apply; Astra Ultrafast on Pro $500 | /plans/codex-pro |
| Business | Per user; see listing | Same estimated ranges as Plus for Standard Business; extend with ChatGPT credits | /plans/codex-business |
| Enterprise and Edu | Contact sales | Flexible pricing: no fixed rate limits, usage scales with credits | /plans/codex-enterprise-edu |
| API key | API pricing | Billed per token; no plan windows | /plans/codex-api-key |
Source for prices and plan contents: Pricing. For every plan side by side with other agents, see /plans.
How to make the allowance last longer
OpenAI’s pricing page lists these steps:
- Control the size of your prompts: precise instructions, no unnecessary context.
- Limit source material to the relevant files, and narrow sources or date ranges.
- Match the output to the need: define audience, format and length.
- Reduce the size of your
AGENTS.md; nest files in the repository to control how much context each task loads. - Limit the number of MCP servers: “Every MCP server adds more context to your messages and uses more of your limit.”
The Codex pricing page adds: “Switch to a lower-cost model for routine tasks.” In the CLI, /model switches models during a session (Models).
