ChatGPT Deep Research Limits by Plan: Free, Go, Plus, Pro, Business

OpenAI does not publish a Deep Research count for ChatGPT Free, Go, Plus or Pro. Its pricing page describes Deep Research as limited on Free and Go, expanded on Plus and maximum on Pro, and the in-product counter shows how many tasks you have left. Where a plan has a fixed monthly allowance, it resets 30 days after your first use, not on the first of the month. On Business and Enterprise credit-based pricing, a Deep Research task costs 50 credits, and ChatGPT Edu includes 5 tasks per rolling 24 hours before credits apply.

ChatGPT Deep Research limits by plan

Source: OpenAI's ChatGPT pricing page, its "Deep research in ChatGPT" help article and its Business, Enterprise and Edu credit rate card, checked October 11, 2026. "Limited", "Expanded" and "Maximum" are OpenAI's labels; it gives no numbers behind them.

How the counter and reset work

ChatGPT tracks Deep Research as tasks. One task is one research run: you describe the outcome, review and edit the proposed research plan, and get a cited report at the end. Your in-product usage counter shows how many tasks remain.

For plans with a fixed monthly allowance, the count resets every 30 days from the date you first used Deep Research. If you first ran it on the 12th, your allowance renews around the 12th of the next cycle, whatever the calendar month. Availability also depends on your country or territory, so a plan that includes Deep Research may not offer it everywhere.

On Edu, the free allowance is a rolling 24-hour window that starts with your first use: five tasks in that window, then credits. On credit-based pricing, each task is billed at 50 credits: Business seats use their included limits first, while Enterprise workspaces with flexible pricing scale usage with credits. Enterprise and Edu admins can switch Deep Research on or off by role.

Deep Research versus search

OpenAI's own rule of thumb is to use search for quick facts and Deep Research for depth. Search pulls recent web information and returns a short answer with links. Deep Research takes longer, reads and analyzes many sources, and can use files you upload, specific sites you choose and connected apps you have enabled, reading from them without making changes. Saving tasks for questions that need a documented report is the simplest way to make a limited allowance last.

What to do when you run out

1. Check the counter. It shows the tasks left and, for monthly allowances, when the 30-day cycle renews.
2. Use search or standard chat for questions that need a quick answer rather than a report.
3. Make each task count. Edit the research plan before it starts, narrow the sources with the site settings, and attach the files that matter.
4. On Business, Enterprise or Edu, ask an admin about workspace credits; each task costs 50 credits on credit-based pricing.
5. Upgrade if you run out every cycle. Plus is described as expanded and Pro as maximum, though OpenAI publishes no count for either.

Don't count on Plus or Pro credits to extend Deep Research. OpenAI lists Codex, ChatGPT Work and its Word, Excel and PowerPoint features as the uses for personal-plan credits, and says credits do not cover every ChatGPT feature.

Which plan fits

• Free or Go to try Deep Research on an occasional question.
• Plus when you run research reports regularly and the limited allowance runs out.
• Pro when Deep Research is part of daily work; OpenAI labels its allowance "maximum".
• Business, Enterprise or Edu when research runs in a company or campus workspace and you want usage to continue on credits.

For how ChatGPT search differs between free and paid plans, see ChatGPT search free vs paid. If you are comparing research tools, Perplexity's limits by plan lists its Research query allowances.