OpenAI Launches Prism, Free AI Workspace for Scientific Collaboration
OpenAI launched Prism in San Francisco on January 27, 2026, releasing a free cloud-based workspace for scientists to draft, edit and collaborate on research documents.
 |
| Credit: OpenAI |
The platform integrates GPT-5.2, the company's model for mathematical and scientific reasoning, directly into LaTeX workflows. Prism allows users with a ChatGPT personal account to create unlimited projects and add unlimited collaborators without subscriptions or seat limits.
Access opened immediately at prism.openai.com, with plans to expand to ChatGPT Business, Enterprise and Education accounts in the coming weeks.
The workspace unifies tools for drafting papers, managing equations, handling citations and coordinating edits among team members.
Scientists can interact with GPT-5.2 inside documents to test hypotheses, revise arguments and search literature from repositories such as arXiv.
Features include converting handwritten diagrams from whiteboards into LaTeX code and making voice-directed changes to text.
OpenAI built Prism on the acquired Crixet platform to embed AI within the full context of research files, including structure, references and figures.
“We’re introducing Prism, a free, AI-native workspace for scientists to write and collaborate on research, powered by GPT‑5.2. Prism offers unlimited projects and collaborators and is available today to anyone with a ChatGPT personal account,” OpenAI said in its release.
Prism receives context from the entire project when users query the model, enabling responses tied to specific equations or sections.
The launch follows increased scientific use of ChatGPT, with 8.4 million weekly messages on advanced topics in fields like mathematics and statistics.
“Software engineering accelerated in part because of amazing models, and in part because of deep workflow integration,” said Kevin Weil, OpenAI's vice president for science.
OpenAI reports examples of GPT-5.2 aiding proofs in statistics and resolving mathematical problems through human-guided applications.
