OpenAI to Expand Academic Program to 200,000 Seats, but Swaps Pro for Plus
OpenAI is doubling its free ChatGPT program for scientists to 200,000 seats, but participants will get the Plus tier instead of Pro. According to the company, researchers did not fully use the capabilities of the more expensive plan. More than 20,000 scientists from 33 countries have already joined the program, and many say that work without AI would take nearly twice as long. Mass distribution of AI to science is happening through a cheaper subscription.
AI-processed from Habr AI; edited by Hamidun News
OpenAI is changing the terms of its ChatGPT for Academic Researchers program. The company is doubling the number of free seats to 200,000, but pays for this with the quality of the tier: instead of ChatGPT Pro, scientists will now get the cheaper ChatGPT Plus. This matters beyond universities. It shows how a major model developer plans to distribute AI to large groups of professional users, and how much it is willing to spend doing so.
What Has Changed
The program launched in July. At that time, 100,000 scientists, mathematicians, and engineers from selected universities around the world could use OpenAI's most powerful models for free through a Pro subscription. The package included research tools and training. Now the participant cap has been raised to 200,000, but the foundation will be Plus.
OpenAI explains the swap simply. According to the company's observations, participants did not use the advanced tier's capabilities enough. The cheaper the subscription, the more people can be reached on the same budget. This is a candid explanation, although it is also clear that Pro costs the company considerably more, and actual demand from researchers turned out lower than expected.
There is an important detail in the numbers. Of the 100,000 seats announced in the summer, just over 20,000 people from 33 countries have joined the program. So the pilot is only partially filled so far, and the new 200,000 remain a plan, not a fact.
What the Pilot Showed
OpenAI studied how more than 1,800 researchers used ChatGPT. Four tasks came up most often:
- reviewing scientific literature;
- analyzing data;
- writing code;
- generating new research ideas.
Respondents said that doing the same work at the same quality level without AI would have taken almost twice as long. The caveat matters: this is participants' self-assessment, not an independent measurement. The company conducted the survey itself, and people who stayed in the program and answered the questions were likely loyal to the tool. Still, the task mix shows where AI has already become routine work: reading papers, processing tables, and writing scripts.
How the New Seats Will Be Allocated
Applications for the next round open on November 23, 2026. OpenAI will offer 50,000 new grants. Each covers the costs of one researcher and up to two co-authors. If all grants are fully used, up to another 150,000 people would get free access to Plus.
The selection mechanism is also changing. In the first round, candidates were chosen by lottery. Now a queue applies: whoever meets the criteria and applies earlier gets a seat. Researchers on the first round's waiting list are promised an email with instructions before applications open. The application map shows interest worldwide: the number of applicants by country is given as of November 12, 2026, and the darker the area, the more applications.
Beyond this program, OpenAI runs two others: ChatGPT Edu for educational institutions and the Rosalind Academic Plan for biology labs at eligible US universities.
What This Means in Practice
For scientists and labs, the takeaway is simple. If you need free access, prepare your application by November 23: with first-come, first-served allocation, speed matters. If you were counting on Pro, for example for heavy computational modes, you will either have to pay for it yourself or work within Plus's capabilities.
For product teams and founders, this is a useful signal. Distributing AI to a professional audience works both as an acquisition channel and as a usage study. OpenAI collects data on scientists' real tasks while getting a generation of researchers used to its tool. Competitors in education and science will likely respond with similar programs.
The limitations are also clear. The near-twofold speedup figure comes from a survey, the number of active participants is five times below the announced ceiling, and the program's terms have already changed within six months. Planning research workflows around a free subscription is risky: the rules could change again.
- Nov 21, 2026Anthropic goes public and raises a record $107.3 billion
- Nov 21, 2026OpenAI launches ChatGPT Wallet: agents buy without ever seeing your card number
- Nov 21, 2026Anthropic goes public and raises a record $81 billion: orders for 7 times more shares than were offered
- Nov 21, 2026OpenAI Launches ChatGPT Wallet: Agents Pay for Purchases Without Seeing Real Card Data
Key Questions Answered
What Has Changed?
The program launched in July. At that time, 100,000 scientists, mathematicians, and engineers from selected universities around the world could use OpenAI's most powerful models for free through a Pro subscription. The package included research tools and training. Now the participant cap has been raised to 200,000, but the foundation will be Plus. OpenAI explains the swap simply. According to the company's observations, participants did not use the advanced tier's capabilities enough.
What the Pilot Showed?
OpenAI studied how more than 1,800 researchers used ChatGPT. Four tasks came up most often: reviewing scientific literature; analyzing data; writing code; generating new research ideas. Respondents said that doing the same work at the same quality level without AI would have taken almost twice as long. The caveat matters: this is participants' self-assessment, not an independent measurement.
How the New Seats Will Be Allocated?
Applications for the next round open on November 23, 2026. OpenAI will offer 50,000 new grants. Each covers the costs of one researcher and up to two co-authors. If all grants are fully used, up to another 150,000 people would get free access to Plus. The selection mechanism is also changing. In the first round, candidates were chosen by lottery. Now a queue applies: whoever meets the criteria and applies earlier gets a seat.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.