Asana cuts model costs 76x in browser tests with GPT-6.1 Sol

OpenAI reported that Asana reduced model costs by 76x in browser tests using GPT-6.1 Sol. The result was presented under the headline “Asana cuts model costs 76x in browser tests with GPT-6.1 Sol.”

Asana is a workplace software company whose products are used to organize projects, tasks and team coordination. The reported outcome places browser-based testing at the center of the cost comparison, with GPT-6.1 Sol identified as the model used in the tests.

A 76x reduction is a relative cost result, indicating that the tested setup used substantially less model spending than the comparison referenced by the report. Lower model costs can be significant for companies evaluating how to deploy AI systems across recurring workplace workflows.

The report was originally attributed to OpenAI. The announcement highlights the role that model selection and testing environments can play in the economics of AI-powered software, particularly for organizations considering broader use of generative AI tools.