OpenAI于7月29日推出「ChatGPT for Academic Researchers」计划，为高校科研人员提供为期一年的免费使用权 [1]。该项目今夏先开放1万个名额，最终目标覆盖10万名研究者 [1]。
参与者可获得包括GPT-5.6 Sol Pro等旗舰模型、专属科研工作区、代码工具Codex与Deep Research等功能，涵盖文献查阅、代码编写、数据分析和论文撰写的完整科研工作流程 [1]。申请资格限定为高校研究教职人员、博士后及相关身份，需通过SheerID机构验证并提交近三年来自arXiv、bioRxiv或ChemRxiv的论文记录 [1]。
项目不包含API接口、无限使用额度与开源模型权重的访问权限 [1]。每周约130万人正在使用ChatGPT进行高等科学和数学研究 [1]。arXiv数学论文的致谢中提及ChatGPT的数量从2月的14篇增至7月头三周的100篇 [1]。
OpenAI unveiled "ChatGPT for Academic Researchers" on July 29, offering 100,000 university researchers one year of complimentary access to its flagship models and research tools [1]. The initiative will roll out 10,000 spots this summer, with the full allocation of 100,000 researchers targeted by 2027 [1].
Eligible participants must hold positions as faculty researchers, postdoctoral fellows, or doctoral candidates at academic institutions, pass institutional verification through SheerID, and demonstrate research publication history within the past three years on platforms such as arXiv, bioRxiv, or ChemRxiv [1]. Through the program, researchers gain access to advanced models including GPT-5.6 Sol Pro, a comprehensive research workspace, the Codex code tool, and Deep Research capabilities, covering the entire research pipeline from literature review to code implementation, data analysis, and manuscript preparation [1].
Approximately 1.3 million users employ ChatGPT weekly for advanced science and mathematics applications [1]. Recent performance benchmarks demonstrate the capabilities available: GPT-5.6 Sol achieved 83 percent accuracy on FrontierMath Tier 4 tasks, compared to 72.5 percent for the previous generation, while Sol Pro resolved 31.5 percent of GeneBench Pro problems [1]. The program excludes API access, unlimited usage tiers, and open-source model weights [1].
