OpenAI is significantly expanding the ChatGPT for Academic Researchers program. It will now offer free access to ChatGPT to as many as 200,000 researchers worldwide. The program launched in July as a pilot project, and applications for the next round will open on Monday, November 23. The company is thus doubling its original target, but it has also changed the conditions under which researchers will receive the tools.
Who can apply
The program is aimed at academics. Researchers from recognized universities and colleges that award academic degrees and are heavily involved in research can apply. Applicants must provide proof of their institutional affiliation and describe what they are working on and how they intend to use artificial intelligence in their research.
What is new is the allocation method. OpenAI will award an additional 50,000 grants, and instead of the previous lottery, they will go to the first applicants who meet the requirements. Each grant will cover one researcher and up to two collaborators, for a total of up to 150,000 people. At the start of the program, each researcher could invite up to four colleagues. So far, 20,000 researchers from 33 countries have joined. Each collaborator must also confirm their institutional affiliation and counts toward the total number of accounts.
What researchers will receive
Here comes the catch. Pilot participants received ChatGPT Pro, OpenAI's highest-tier plan, while the new grants offer the less expensive Plus subscription. The company says that most researchers in the pilot would have stayed within the limits of the Plus plan and that this change will allow it to reach more teams.
Researchers have access to ChatGPT, ChatGPT Work, and the Codex programming tool. Codex will help them write and debug code, analyze data, and create reproducible workflows. ChatGPT Work can handle longer projects, such as finding research funding, preparing grant applications, reviewing academic literature, or writing manuscripts. Biologists can use more than 75 specialized tools in genetics, genomics, sequencing, protein modeling, and drug development. However, access to the specialized GPT-Rosalind model requires approval through a separate review process.
The program also includes support. OpenAI is preparing training for different levels of experience, from the basics and streamlining workflows to advanced applications in research. Participants will receive help from experts who understand scientific work and can advise them on how to integrate the tools into their daily practice.
What OpenAI is up to
The program is a source of valuable information for OpenAI. Feedback from researchers is intended to show the company where its models help, where they fail, and how to improve them to make them more useful to the research community. OpenAI also believes that science advances faster when more researchers have tools to develop their own ideas. The program is part of a broader plan under which the company intends to invest more than $250 million in supporting external scientific research by 2027.
OpenAI already has its first results. It published a study on how more than 1,800 researchers used ChatGPT during the pilot. Reading and studying accounted for an average of 28 percent of their research queries, while writing accounted for 27 percent. Respondents said that without artificial intelligence, the same work would have taken them almost twice as long. However, this figure comes from the participants themselves, and OpenAI acknowledges that the study does not prove that AI accelerates scientific discoveries. It cites a lack of funding, personnel, and laboratory capacity as obstacles to testing new ideas.
OpenAI is not the only company vying for researchers' attention. Anthropic launched Claude Science, an AI-powered workspace for laboratories, in June, and Google DeepMind made its scientific models available to 200 European researchers in October.



