OpenAI Opens 50,000 ChatGPT Research Grants November 23 | Zeniteq
Loading
Zeniteq is now available on mobile for iPhone, iPad and Apple Watch.Download nowLearn more
OpenAI Opens 50,000 ChatGPT Research Grants November 23
Eligible researchers can apply for free ChatGPT Plus access for themselves and up to two collaborators, with awards moving from a lottery to first-come, first-served selection.
OpenAI will open applications for 50,000 new academic research grants on November 23, 2026, shifting from a lottery to a first-come, first-served model. Each grant provides ChatGPT Plus access for the recipient and two collaborators, potentially supporting 150,000 researchers. While this expansion offers broader access, applicants should note the downgrade from the pilot's Pro tier to Plus, which imposes stricter usage limits. Researchers must verify if these limits accommodate their specific workflows, as the grant excludes data from model training but does not address broader institutional privacy or data retention requirements.
OpenAI will open applications for 50,000 new ChatGPT for Academic Researchers grants on Monday, November 23, 2026. The company announced the expansion on November 16, setting a broader goal of providing free ChatGPT access to as many as 200,000 researchers worldwide.
Two things changed in this round, and applicants should understand them before applying. Awards will go to the first 50,000 applicants who meet the eligibility requirements, replacing the pilot’s lottery. New recipients will receive ChatGPT Plus, while the summer pilot provided ChatGPT Pro.
Each grant covers its recipient and up to two collaborators. That gives this round a maximum reach of 150,000 researchers, although the total depends on how many collaborator places recipients actually use.
What Changes From the Researcher Pilot
OpenAI launched the pilot in July and says 20,000 researchers across 33 countries have joined since then. Its November announcement describes the next round as an expansion informed by participant feedback and a review of usage.
Detail
July pilot
November application round
Selection method
Lottery-based
First 50,000 qualifying applications
Included ChatGPT plan
Pro
Plus
New grant allocation
Not specified in the expansion announcement
50,000 grants
People covered per new grant
Not specified for the pilot in this announcement
Recipient plus up to two collaborators
OpenAI says it changed the selection process following researcher feedback. It also says most pilot participants would have stayed within ChatGPT Plus usage limits, which informed its decision to offer Plus for new awards. Participants will receive access to OpenAI’s frontier models and training on using ChatGPT in research, according to the official announcement.
The distinction between grants and people matters here. 50,000 grants are not 50,000 individual places, because each award can cover three people. Equally, the 200,000-researcher figure is the program’s overall goal, not the number of new grants available on November 23.
At full uptake, the next round’s 150,000 places would equal 75% of that goal. That is capacity calculated from OpenAI’s published figures, not a forecast of how many researchers will activate access.
How to Prepare for November 23
OpenAI’s announcement establishes the application date and selection principle, but it does not enumerate every eligibility restriction or country where applicants can participate. Researchers should use the program information linked from the official announcement to check the current requirements.
A practical preparation list:
Check whether you are already on the waitlist. OpenAI says qualifying waitlisted researchers will receive an email before applications open, with instructions for activating their accounts.
Confirm eligibility and country availability. The worldwide target should not be treated as confirmation that every researcher in every location qualifies.
Identify up to two potential collaborators. Each new award can cover that small team, although the announcement does not explain the collaborator invitation process.
Look for opening-time and application details. The announcement gives November 23 as the opening date but does not specify an opening hour or time zone.
Check award duration and account terms. The post does not state how long the new Plus access lasts or explain how it interacts with an existing paid subscription.
I’d treat this as a scheduled application task, not an opportunity to revisit whenever convenient. A lottery allows applicants to enter during a window without necessarily benefiting from being early. First-come, first-served selection makes application order consequential, subject to eligibility.
That change can also make preparation and time-zone clarity more important. The published announcement does not explain how incomplete applications, eligibility checks or tied submission times will be handled, so applicants should avoid assuming rules beyond the first-come commitment.
Plus Access Broadens Reach, With a Workload Tradeoff
OpenAI’s stated reasoning for switching plans is straightforward: most pilot participants would have fit within Plus usage limits, so the company is offering Plus to extend access to more research teams.
That supports a useful but limited conclusion. Plus may cover the usage patterns of most pilot participants; it is not a promise that every research workload will fit. The announcement does not give a percentage for “most,” publish the relevant usage distribution or identify the model-specific limits used in its comparison.
For a researcher evaluating the offer, the relevant question is whether Plus can support the intended workflow. A project involving occasional literature questions and code assistance has different access needs from one involving sustained, repeated model use. The announcement’s aggregate finding cannot resolve that choice for an individual lab.
My view is that the switch is best understood as a breadth-versus-capacity decision. OpenAI is using a lower access tier to reach more people, based on its own usage review. That can be valuable for teams without an AI subscription budget, while leaving heavier users with a reason to inspect the current plan limits carefully.
The change applies to new grants. The announcement does not say that existing pilot participants are being moved from Pro to Plus.
The Privacy Promise Concerns Training, Not Data Collection
OpenAI states: “By default, we don’t use data shared through the program to train our models.” That is the relevant program-specific commitment for researchers assessing how their inputs may be used.
That wording deserves careful reading. Semafor’s coverage describes the company as saying it will not collect data from enrolled users. The official announcement makes a narrower statement about model training by default. Those are different claims, and researchers should rely on the official wording when making a data-handling decision.
Excluding data from model training does not, by itself, answer questions about retention, access controls, storage location or permission to upload sensitive material. The expansion announcement does not provide those details.
For work involving patient information, unpublished findings or restricted datasets, I’d separate the subscription decision from the data-approval decision. Free access can remove a cost barrier; it cannot replace institutional requirements or the terms governing a particular dataset.
What the Pilot’s Productivity Evidence Actually Shows
OpenAI also published research about how scientists used ChatGPT during the pilot. Its announcement says surveyed participants estimated that producing the same research work at the same quality would have taken nearly twice as long without AI.
The Boston Report’s account of the company’s study provides more detail: OpenAI analyzed 580,186 scientific messages from 1,819 scientists and surveyed 251 participants. The reported estimate was that the work would have taken 1.9 times as long without AI.
That is a participant estimate, not a measured speedup from a controlled experiment. The Boston Report notes that OpenAI’s researchers acknowledged the findings do not demonstrate causality.
The study is useful evidence about how participating researchers perceived the tool’s value. It offers less certainty about how much time a new recipient will save, whether output quality will remain consistent or whether faster individual tasks will translate into faster scientific discoveries. Those distinctions matter when deciding how much of a research workflow to build around the grant.
Other Research Access Routes Serve Different Needs
OpenAI’s announcement also points to alternatives that may better fit an institution or a larger lab:
ChatGPT Edu lets universities provide access for research, teaching and learning, with administrative controls for managing access.
The Rosalind Academic Plan offers eligible biology labs at U.S. universities up to 30 seats at no cost for up to 12 months. Access to GPT-Rosalind requires approval through OpenAI’s trusted-access process.
Those conditions come from the same research-access announcement. The Rosalind offer’s duration and seat count should not be carried over to the new ChatGPT Plus grants.
For a researcher and two collaborators, the November round offers a clearly defined team size. For a department needing centrally managed access, or a qualifying biology lab needing specialized tools, the other routes deserve a separate look before choosing an access plan.
Frequently Asked Questions
4 questions
1
When Do Applications Open?
Applications open Monday, November 23, 2026. OpenAI’s announcement does not specify an opening hour or time zone.
2
Do New Research Grants Include ChatGPT Pro?
No. New grants provide ChatGPT Plus. The earlier pilot provided Pro, and OpenAI says most pilot participants would have stayed within Plus usage limits.
3
How Many People Can One Grant Cover?
One grant covers the recipient and up to two collaborators, for a maximum of three researchers. Across 50,000 grants, that creates capacity for up to 150,000 people.
4
What Should Waitlisted Researchers Do?
Check for an email from OpenAI before applications open. The company says qualifying waitlisted researchers will receive instructions for activating their accounts.