OpenAI announced on July 29, 2026 that it will give 100,000 researchers at academic institutions free access to its frontier AI models through a new program called ChatGPT for Academic Researchers. The company frames the effort as an attempt to speed up scientific discovery by putting its most capable models — normally reserved for its $200-a-month ChatGPT Pro tier — directly into the hands of working scientists rather than only well-funded labs and corporate research teams.

What researchers get

Each accepted participant receives access equivalent to a ChatGPT Pro subscription at no cost, including higher usage limits, expanded deep research capability, and larger context windows, with access to OpenAI's frontier models such as GPT-5.6 Sol Pro at launch. Approved researchers can also invite up to four collaborators from their institution, extending access beyond a single principal investigator. The program targets researchers in biology, chemistry, computer science, engineering, mathematics, and physics at recognized, degree-granting institutions with a high level of research activity; OpenAI says model weights themselves remain off-limits, with access delivered strictly through its hosted products.

Rollout timeline

The first cohort of 10,000 researchers is set to begin receiving access this summer, with OpenAI planning to expand the program to the full 100,000-researcher target on a rolling basis through 2027. Scientists at Princeton's Institute for Advanced Study and France's École normale supérieure are among the institutions already onboarded. OpenAI has bundled the program into a wider commitment of more than $250 million through 2027 for external scientific research.

Why it matters

The move continues a pattern among frontier labs of subsidizing access for academia as a way to build goodwill, generate research findings that showcase model capabilities, and seed a pipeline of researchers trained on their specific tools. It also arrives as OpenAI faces growing scrutiny over compute costs and its path to profitability, making a large free-access commitment notable timing: the company is effectively betting that donated inference capacity for science produces enough reputational and ecosystem return to justify the cost, at a moment when compute itself is its scarcest and most expensive resource.