OpenAI grants 100,000 researchers free ChatGPT access for research

Thursday, 30 July 2026 at 06:00

OpenAI geeft 100.000 wetenschappers gratis toegang tot ChatGPT voor onderzoek

OpenAI is launching a new global program that will ultimately give 100,000 academic researchers free access to its most advanced ChatGPT models. The initiative aims to accelerate scientific breakthroughs and is part of an investment of more than $250 million in scientific research through 2027.

The first 10,000 researchers will get access this summer. Early participating institutions include the Institute for Advanced Study (IAS) and École normale supérieure (ENS). OpenAI plans to scale the program to 100,000 researchers worldwide.

OpenAI

@OpenAI

·Follow

We’re giving scientists, mathematicians, and engineers free access to our frontier models—starting with 10,000 researchers and expanding to 100,000 through 2027. ChatGPT for Academic Researchers is built to accelerate discovery across disciplines.

Watch on X

5:19 PM · Jul 29, 2026

7.1K

Read 464 replies

Free GPT-5.6 access for universities

With its new ChatGPT for Academic Researchers program, OpenAI wants to give scientists at universities and research institutes access to its latest AI models, including the GPT-5.6 family.

Participants can use GPT-5.6 Terra, Luna, and the most powerful model, GPT-5.6 Sol Pro. They also get access to Codex, ChatGPT Work, higher usage limits, larger context windows, and specialized AI tools for scientific research.

Researchers may invite up to four colleagues from the same institution. According to OpenAI, data within these workspaces is not used to train its models by default, and enterprise-grade security and privacy protections apply.

AI for the entire research workflow

OpenAI no longer frames ChatGPT as just a chatbot, but as a broad research platform.

Scientists can use the software for a wide range of tasks, including:

literature reviews;
hypothesis generation;
genetic analyses;
protein modeling;
data analysis;
coding via Codex;
grant applications;
scientific publishing.

The platform also supports more than 75 specialized life-science tools for genetics, sequencing, single-cell analysis, protein studies, and drug development. Through connectors, researchers can integrate with GitHub, Hugging Face, Google Drive, Databricks, Zotero, and LaTeX.

AI adoption in science is surging

According to OpenAI, roughly 1.3 million people now use ChatGPT for advanced scientific and mathematical problems, sending about 8.4 million messages each week.

Usage is rising fastest in mathematics. OpenAI points to a growing number of arXiv papers explicitly crediting ChatGPT in the research process. The company says mentions climbed from 14 papers in February to 100 in the first three weeks of July 2026.

The nature of the work is shifting too. Researchers deeply invested in AI increasingly hand off complex tasks that would normally take hours of human effort. OpenAI says this group is nearly twice as likely to use AI for tasks lasting four hours or more.

Stronger scores on scientific benchmarks

OpenAI stresses that GPT-5.6 Sol Pro is designed for complex scientific workloads.

The company reports an 83 percent score on FrontierMath Tier 4, a research-level math benchmark—up from GPT-5.5’s 72.5 percent on the same test.

On GeneBench Pro, a benchmark for biological data analysis and scientific reasoning, GPT-5.6 Sol Pro also performs better, successfully completing 31.5 percent of complex research tasks, according to OpenAI.

Training and hands-on support included

Beyond free model access, OpenAI will provide training and personalized support.

Researchers can get guidance on integrating AI into day-to-day work. OpenAI also aims to build a community where scientists share experiences and offer feedback on model usage.

The company says that feedback will help tailor future AI systems to the needs of the scientific community.

Who can apply?

The program is open to researchers affiliated with accredited universities and research institutions with strong research activity.

Applicants must verify their academic affiliation and specify how they plan to use AI. Once approved, they can invite up to four colleagues, who must also confirm affiliation with the same institution.

OpenAI expects to scale the program to 100,000 researchers by the end of 2027.

Why it matters

With this move, OpenAI is steering generative AI decisively toward scientific research. Where AI was first used mainly for text generation and software development, the company now positions its latest models as research co-pilots for complex problems in physics, biology, medicine, mathematics, and engineering.

Free access for a large academic cohort could further accelerate AI adoption across universities. At the same time, OpenAI gains valuable insight into how scientists actually use the technology—feedback that can shape the next generation of models.

You can submit an application here, by the way.