OpenAI Opens ChatGPT for Researchers to 200,000 Scientists — and Google Adds Gemini 4 Carbon
OpenAI is doubling its free ChatGPT programme for academic researchers to reach up to 200,000 scientists worldwide, with 50,000 new grants opening for applications on 23 November 2026. On the same day, Google DeepMind added free Gemini 4 Carbon to its scientist access programme for 10,000 scientists. Two of the biggest AI labs made the same bet on the same day: give researchers their best models for free.
ChatGPT for Academic Researchers is OpenAI's programme giving scientists free ChatGPT Plus for research work. It launched as a pilot in July 2026 with a target of 100,000 researchers. The new goal of 200,000 doubles that target, and the next round changes how grants are handed out: first come, first served, instead of by lottery.
What OpenAI actually announced
OpenAI announced the expansion on 16 November 2026, alongside a study of how pilot researchers actually used ChatGPT. The pilot has enrolled 20,000 researchers across 33 countries since July. The next round opens on Monday 23 November, offering 50,000 new grants. Each grant covers the recipient and up to two collaborators, so this round alone could support up to 150,000 researchers with free ChatGPT Plus.
There is one change researchers should not miss. Pilot participants received ChatGPT Pro, OpenAI's top tier. New grants come with ChatGPT Plus instead. OpenAI's explanation is practical: reviewing pilot usage, it found most participants stayed within Plus limits, so switching tiers lets the programme reach far more people. The grants last up to 12 months, and data shared through the programme is not used to train OpenAI's models by default.
Eligibility is narrower than the headline suggests. According to OpenAI's help centre, applicants must be research faculty or postdoctoral researchers at an eligible academic institution in a supported country, have published relevant research in the past three years, and plan to use ChatGPT for academic research. Eligible fields are biological sciences, chemistry and materials, computer science, earth and planetary sciences, engineering, mathematics and physics. Applications are open worldwide but exclude China and Hong Kong, in line with OpenAI's regional availability policies.
![]() |
| Credit: The Motley Fool |
The numbers that matter
The two announcements are easy to confuse, so keep the figures separate. OpenAI's programme is the bigger one by headcount; Google's is the smaller, more selective one. Here is the split:
| Programme | Who gets in | What they get | When |
|---|---|---|---|
| OpenAI ChatGPT for Academic Researchers | Up to 200,000 researchers (goal) | Free ChatGPT Plus for 12 months | 50,000 grants open 23 Nov 2026 |
| Google DeepMind Antigravity scientist programme | 10,000 scientists | Six months of free Gemini 4 Carbon access | Applications open now, rolling weekly selection |
| OpenAI pilot so far | 20,000 researchers in 33 countries | ChatGPT Pro (pilot tier) | July to November 2026 |
The 200,000 figure is a programme goal, not current enrolment. That distinction matters: today the programme has 20,000 researchers, and the November round could add up to 150,000 more people if every grant covers three researchers. OpenAI has committed more than $250 million through 2027 to support external scientific research, and this programme sits inside that commitment.
What the study actually found
The expansion arrived with a study meant to justify it. OpenAI's Economic Research team published Science at the Frontier: How Scientists Use ChatGPT, analysing 580,186 scientific messages from 1,819 pilot researchers between 20 August and 22 September 2026, alongside a survey of 251 participants conducted from 30 September to 7 October.
The headline finding is striking: surveyed scientists estimated that the same work, at the same quality, would have taken nearly twice as long without AI — about 1.9 times as long. Most of the usage was unglamorous and practical. Around 68% of researchers used ChatGPT to develop research ideas, and the bulk of messages went to drafting, coding help, data analysis and literature work. End-to-end autonomous research — asking the AI to coordinate literature review, ideation and downstream analysis — accounted for only 0.4% of the average scientist's scientific messages.
OpenAI is unusually direct about what those numbers do not prove. The participants were a selected group who applied for expanded AI access and chose to answer an AI survey, so the results may not generalise to all scientists. The findings do not establish causal effects either — the study documents how AI fits into research work, not whether it produces discoveries faster.
![]() |
| Credit: Nova Southeastern University |
Google's half: Gemini 4 Carbon for 10,000 scientists
Google DeepMind added free Gemini 4 Carbon to its Antigravity scientist programme for 10,000 scientists, with six months of free access on Antigravity's highest usage tier. Thousands of scientists had already joined the programme, which initially offered Gemini 4 Argon; current and future participants now receive access to Carbon as well, with applications remaining open and researchers selected on a weekly rolling basis.
Carbon is Google's strongest Gemini 4 model, built for demanding tasks involving extended reasoning, verification and work across multiple steps, including open-ended scientific investigation. It became generally available to paying developers and enterprises on 10 November 2026 through the Gemini API, and to Google AI Ultra subscribers in Antigravity, the Gemini app and AI Mode in Search. For paying customers it costs $2 per million input tokens and $10 per million output tokens at the introductory rate, doubling to $4 and $20 after 30 November 2026 — which makes the six months of free access genuinely valuable for researchers running long jobs.
Google says Carbon powered AlphaProof Code agents that proved the security of Signal's cryptographic code in weeks rather than years, spotted wet-lab errors straight from lab notebooks, and found 40 times more related protein sequences for hard-to-predict proteins. Those are Google's own claims about its model, not independent results, and they should be read that way.
Who should apply, and what to watch
If you are a research faculty member or postdoctoral researcher with a publication in the past three years, the OpenAI round is the one to prepare for now. Applications open on 23 November and grants go first come, first served, so eligible researchers should have their details ready rather than waiting for a reminder. The first-come-first-served system replaces the pilot's lottery, which OpenAI changed after researchers asked for a clearer process.
Google's route is different: apply through Antigravity for six months of access, with weekly rolling selection. The two can be combined — OpenAI's grant is a general ChatGPT Plus seat for everyday research; Google's is its strongest model for long, multi-step jobs.
![]() |
| Credit: Arttia |
Why both labs moved on the same day
The timing is not a coincidence. Frontier models are now genuinely useful for scientific work, and both labs want their models to be the ones scientists build workflows around. For readers of the State of AI Report 2026, this is the next chapter; the earlier Gemini 4 Carbon launch explains the model, the Claude AI drug discovery work shows the same pattern in another lab, and the AI breakthrough complexity theory claim is a reminder that bold research claims need careful reading.
Key Takeaways
- OpenAI is expanding ChatGPT for Academic Researchers to a goal of 200,000 scientists; 50,000 new grants open 23 November 2026, first come, first served.
- New grants give free ChatGPT Plus (not Pro) for up to 12 months, covering the recipient and up to two collaborators — up to 150,000 people this round.
- The pilot has 20,000 researchers in 33 countries; surveyed pilot users estimated the same work would have taken about 1.9 times as long without AI — a self-reported estimate, not a measured gain.
- Google DeepMind separately added free Gemini 4 Carbon to its Antigravity programme for 10,000 scientists, with six months of free access and weekly rolling selection.
- Eligibility is limited to research faculty and postdoctoral researchers with a recent publication; China and Hong Kong are excluded from OpenAI's programme.
Conclusion
Two labs, two programmes, one message: the best AI models are now free for the scientists who can use them best. OpenAI's 200,000-researcher goal and Google's 10,000-scientist Carbon offer are both real, both time-limited, and both worth applying for if you qualify. But keep the numbers separate, read the eligibility rules before 23 November, and treat the productivity claims as what they are — promising, self-reported, and still unproven at scale. The models are free; the judgement about how to use them still belongs to the researcher.
Sources
- OpenAI — "Broadening access to AI for scientists" (announcement, 16 November 2026)
- OpenAI — "Science at the Frontier: How Scientists Use ChatGPT" (Economic Research team study, November 2026)
- Google DeepMind — Gemini 4 Carbon scientist access programme via Google Antigravity (announced 16 November 2026)
- Google DeepMind — "Gemini 4 Carbon is now available" (10 November 2026)
- OpenAI Help Centre — ChatGPT for Academic Researchers eligibility requirements



Comments
Post a Comment