A Reddit user ran gpt-oss-120b, OpenAI's largest open model, cutting it between six devices, including a Samsung Galaxy S24+. Mounting works, but it only generates 1.1 tokens per second.
This story is only covered by news sources that have yet to be evaluated by the independent media monitoring agencies we use to assess the quality and reliability of news outlets on our platform. Learn more here.
A Reddit user ran gpt-oss-120b, OpenAI's largest open model, cutting it between six devices, including a Samsung Galaxy S24+. Mounting works, but it only generates 1.1 tokens per second.