Announcing Four New AI Models on the Clarity Platform

Claude Sonnet 4, o3, Gemini 2.5 Flash, and Gemini 2.5 Pro are now available on the Clarity platform, offering Clarity users a more up-to-date version of Claude Sonnet, a reasoning-focused model from OpenAI, and the platform’s first models from Google. 

Details

The latest additions to Clarity include two models from Google’s Gemini family, hosted on Yale’s Google Cloud Platform subscription. Gemini 2.5 Flash is a lightweight yet powerful option for quick and efficient responses, while Gemini 2.5 Pro targets tasks that require deeper reasoning and complex problem-solving capabilities. Both Google models have large context windows, allowing them to work with large amounts of information included in the message prompt text. 

Claude Sonnet 4 is the latest model from Anthropic, providing Clarity users with an updated version of a model that many prefer for its writing quality. 

Finally, the o3 model represents OpenAI’s newest reasoning-focused technology, which responds more slowly than the company’s other models, such as GPT-4o and GPT-4o mini, but excels at handling complex analytical tasks. Use o3 for thinking through tricky problems, GPT-4o for day-to-day tasks, and GPT-4o mini for quick responses and lower energy consumption.

Current Limitations and Security

For now, the new agents handle only text-based conversations. They do not support file uploads or advanced tools such as file search, code execution, data analysis, file creation, or image processing capabilities.

Like other agents on the Clarity platform, these new AI assistants are approved for use with low, moderate, and most high-risk data classifications, with the notable exception of electronic Protected Health Information (ePHI), which remains restricted.

Clarity dropdown menu showing Claude Sonnet 4 and Gemini and other agents

How to Access

Clarity users can find the new agents in the platform’s agent dropdown menu at clarity.yale.edu. We have also updated the format of agent descriptions in Clarity to provide more detailed information about each model’s capabilities and knowledge cutoff dates when starting new conversations.