OpenAI expands free researcher access, switches grants to Plus | Magica
OpenAI targets 200,000 researchers with free ChatGPT, switching new grants to Plus
Editorial Team
••📖5 min read
OpenAI will open 50,000 new ChatGPT access grants on November 23, covering up to 150,000 researchers and collaborators. New recipients get Plus instead of Pro, with first-come selection replacing a lottery. Its pilot study reports perceived time savings but does not establish faster discovery.
OpenAI will offer 50,000 new grants, each covering a researcher and up to two collaborators.
New recipients get ChatGPT Plus instead of Pro; OpenAI says most pilot users would have stayed within Plus limits.
The program’s 200,000-person goal is distinct from its reported enrollment of 20,000 researchers in 33 countries.
OpenAI, the developer of the ChatGPT AI assistant, will open another round of free academic research access on November 23, expanding a program launched in July while lowering the subscription tier for new recipients. Its November 16 announcement offers 50,000 grants on a first-come, first-served basis to eligible applicants, replacing lottery selection.
The trade-off is broader access with lower usage allowances. Pilot participants received ChatGPT Pro; new grants provide the less expensive Plus subscription. OpenAI says its review found most pilot participants would have stayed within Plus limits, allowing it to extend access to more teams. It does not disclose how many exceeded those limits or the cost of the expansion.
Company-reported waitlist signups as of November 12, 2026, shown on a logarithmic scale. Signups are distinct from enrollment. Graphic: OpenAI. Source: OpenAI.
OpenAI’s company-reported access goal increased from 100,000 to 200,000 researchers. These targets are distinct from enrollment and the next round’s capacity. Chart: Magica. Source: OpenAI program launch.
What the expanded grants cover
ChatGPT for Academic Researchers supplies model access, research tools and training. The July 29 launch aimed to reach 100,000 researchers through 2027 and allowed each recipient to invite up to four collaborators from the same institution. The new round reduces that allowance to two, making its maximum capacity 150,000 people: 50,000 recipients plus 100,000 collaborators.
South Korea is strengthening device checks and considering public Wi-Fi shutdowns before its November 19 college entrance exam, after a tax-accountant exam top scorer admitted using AI to cheat. The response still depends heavily on human screening, with no electronic device detectors planned for this year's test.
Editorial Team
OpenAI has doubled the program’s overall goal to 200,000 researchers. That target and the next round’s maximum capacity describe potential access, not current enrollment. The company says 20,000 researchers in 33 countries have joined since July.
Institutional eligibility remains a gate alongside application timing. At launch, OpenAI restricted the program to qualifying researchers at selected institutions—recognized, degree-granting colleges or universities with high research activity. Applicants had to verify their affiliation and describe their active research and intended scientific use; collaborators also had to verify affiliation and counted toward the account total.
The expansion still requires applicants to qualify. OpenAI says eligible researchers already on the waitlist will receive activation instructions before applications open. The announcement does not specify the duration of the new grants or a numerical schedule of Plus usage limits.
Company-reported survey estimate: equivalent work would take about 1.9 hours without AI for one hour with AI available. This does not establish causal effects or faster discovery. Chart: Magica; source: OpenAI. Source: OpenAI pilot research.
Time-saving estimates, not measured discoveries
OpenAI’s Economic Research team examined 580,186 scientific messages from 1,819 program scientists between August 20 and September 22, 2026. Its accompanying study also analyzed a September 30–October 7 survey with 251 participants, with response counts varying by question.
Across eight research activities, respondents estimated that an hour of work with AI available would have required about 1.9 hours without it to achieve the same quality. Each activity received equal weight in that average, with 249–251 respondents per activity. When someone reported no AI use for an activity, the calculation treated the time as unchanged.
This is a self-reported comparison, not an observed near-doubling of productivity or a measured acceleration in discovery. The study explicitly says its findings do not establish causal effects. OpenAI also identifies staffing, funding and experimental capacity as constraints on scientists’ ability to test ideas.
The usage evidence reaches beyond administrative assistance: 68% of the scientists used ChatGPT to develop research ideas, defined as sending at least three messages about ideation during the study. OpenAI describes pilot researchers exploring links between laboratory and patient data for tuberculosis drug development and turning museum-specimen photographs into searchable databases. Those examples illustrate workflows; they do not establish successful drug development or a general improvement in research outcomes.
The data policy also requires precision. OpenAI says program data is not used to train its models by default. A report describing that assurance instead says the company will not collect participants’ data, a broader claim than the announcement supports. The study says automated classifiers categorized messages, while the research team saw those categories without reading underlying messages or responses.
Research access has several competing routes
The July launch placed the program inside OpenAI’s commitment of more than $250 million through 2027 for external scientific research and discovery. That broader commitment is not a disclosed budget for this access round. Neither the announcement nor its usage rationale establishes OpenAI’s cost per researcher.
OpenAI’s expansion announcement also describes ChatGPT Edu, which gives university communities access for research, teaching and learning with institutional administrative controls. For biology, its specialized GPT-Rosalind model connects biological reasoning with scientific tools and databases. The Rosalind Academic Plan offers eligible U.S. university labs up to 30 free seats for up to 12 months, subject to trusted-access approval. These are distinct access arrangements, not interchangeable grant terms.
OpenAI also says it contributed $5 million to a Simons Foundation initiative helping mathematicians examine and build on AI-generated results. That effort addresses checking and understanding outputs as well as access to tools.
Other AI developers are pursuing scientific work too. A report on the expansion describes Anthropic’s June launch of Claude Science, an AI workbench for laboratories, and Google DeepMind’s October opening of science models to 200 European researchers. Their different scope means those figures cannot establish which provider offers better value or wider effective research capacity.
What the November round needs to establish
For applicants, the next step is qualification and activation through the waitlist or the November 23 opening. For OpenAI, the unresolved question is whether Plus accommodates scientific workloads as the program expands beyond its pilot population.
The company says participant feedback will shape future rounds. Reporting how often recipients hit usage limits would test its rationale for the subscription switch. Evidence that ideas were validated and research outcomes improved—alongside the funding, staff and experimental resources required—would address the larger claim that wider AI access accelerates discovery.
Russia’s objections to observer participation stalled the opening of a Geneva weapons conference, where governments must decide whether to authorize negotiations on autonomous weapons and whether any resulting rules should be binding.
Apple’s iOS 27.3 beta ties higher allowances for server-based AI to existing iCloud+ plans. Basic Siri commands remain free, but undisclosed daily limits leave the value of paid access uncertain.
The partners plan to combine IBM quantum processors, NVIDIA computing and Fermilab physics expertise to study modular quantum systems. A separate decoder release promises faster error-correction research, with important limits on its performance comparisons.
Federal Reserve Chairman Kevin Warsh outlined AI scenarios spanning shared prosperity, job displacement and an investment downturn. The exercise offers no forecast or rate guidance as colleagues worry that the infrastructure buildout is adding to inflation.
OpenAI is offering approved organizations a free classifier for suspected child sexual abuse images, including AI-generated material. Early results suggest less review work, but existing detection alternatives and the need for human review qualify the launch.
Nvidia plans scientific computing racks for early 2027, extending Vera beyond its initial AI deployments. Its 2.4× performance-per-watt claim comes from one airflow simulation against AMD Turin; performance against newer rivals and delivered costs remain unresolved.
OpenAI’s CEO expressed empathy for Nvidia founder Jensen Huang but withheld judgment on reinstatement. His confidence in continued AI financing comes with a direct commercial stake and unresolved deployment risks.
Google’s 8,192-chip Ironwood submission completed the MLPerf DeepSeek-V3 benchmark in 1.70 minutes, close to NVIDIA’s 1.53 minutes at the same accelerator count. Google’s scaling data also shows the extra chip-time required for its fastest result, leaving cost and energy comparisons unresolved.
TensorWave will supply dedicated AMD infrastructure under a six-year OpenAI agreement, with delivery expected from 2027. Its operating record and financial backing sit alongside an undisclosed contract price and a recently terminated site lease.
California alleges OpenAI endangered the public and misled consumers by testing powerful AI models without adequate controls. The lawsuit follows a Hugging Face intrusion that OpenAI’s own account traced to containment failures, missing monitors and mishandled warning signs.
Perplexity has launched Frontier Research to develop AI that can formulate questions, test hypotheses and produce interpretable discoveries. The company promises more than 10,000 GPUs, but existing automated-research systems show why computing resources and generated papers alone cannot establish scientific value.
The new lab led by Stanford professors Yejin Choi and Carlos Guestrin has backing from Nvidia, Cisco and Snowflake. It promises systems that improve on company work, but research benchmarks leave production costs and customer control unproven.
NVIDIA announced a Rubin HPC GPU with more than 50,000 CUDA cores, sparse data acceleration and seventh-generation NVLink. The design targets simulations used by scientists and AI agents, but missing performance, power and pricing details leave its practical advantage unproven.
OpenAI reports production healthcare workflows at CVS and Cigna and stronger computer-use evaluations. The results show progress across disconnected software, but leave complete-task reliability and total deployment costs unresolved.
President Donald Trump supported a return for Nvidia co-founder Jensen Huang, but the company has installed interim leadership and begun a successor search amid scrutiny of his conduct and alleged Chinese intelligence ties.
Intel disclosed the bandwidth target for its next Xeon server processors at SC26, alongside previously announced core and cache limits. Application performance, power requirements and launch timing remain unresolved.
Mercedes-Benz plans California early access to city-driving and hands-free highway assistance in December, ahead of a 2027 U.S. rollout. Mixed road tests and unannounced subscription prices leave buyers waiting for evidence of everyday value.
Researchers estimate that 403 large U.S. data centers consumed 81.8 TWh in a year. Their regional-grid analysis reveals substantial fossil-fuel dependence, but cannot isolate AI demand or fully expose the underlying facility data.
HSBC lowered its Zhipu share-price target to HK$500 and delayed forecast adjusted breakeven by two years. Improving API margins and substantial fundraising cushion a business whose gross profit still covers little of its research bill.