OpenAI has released GPT-6.1 Terra, an updated model designed for routine office tasks such as screen navigation, application interaction, and drafting responses. The company reports that the new model handles computer-based operations with increased speed and accuracy compared to its predecessor, GPT-5.6 Terra.
The standard API pricing for GPT-6.1 Terra is 50% lower than that of the GPT-5.6 Terra model. The update is currently available to Plus, Pro, Enterprise, and Edu users within the ChatGPT Work and Codex platforms. Free users continue to access the lower-tier GPT-5.6 Luna model.
Internal testing shows a reduction in factual error rates for GPT-6.1 Terra. At low reasoning effort, errors dropped from 53.6% to 16.7%, and at maximum effort, they decreased to 4.3%. On the Terminal-Bench 4.0 benchmark, the model successfully completed 50.3% of tasks, compared to 21.5% for the previous version.
OpenAI claims that GPT-6.1 Terra completes tasks more than three times faster on average than the GPT-6 Astra and GPT-6.1 Sol models. While Terra is optimized for routine office work, the company recommends using GPT-6.1 Sol or GPT-6 Astra for complex coding or in-depth research.
