OpenAI has rolled out GPT-6.1 Terra, a new AI model aimed at routine professional tasks. The company says it is quicker and cheaper to run than its predecessors. It calls the model a 'significant step up' from GPT-5.6 Terra, with gains in accuracy, coding and operating desktop applications.
Built for visual and computer-based tasks
OpenAI says the model brings together image understanding, the ability to operate software and more natural writing. Tasks it is meant to handle include drafting proposals, interpreting technical drawings, completing forms and editing video. Among the professional tests the company ran, GPT-6.1 Terra built 3D models from architectural drawings and reviewed animated films to produce time-coded reports.
Benchmark claims
On the Agents' Last Exam benchmark, OpenAI says GPT-6.1 Terra scored 15 percentage points above GPT-5.6 Terra at medium reasoning effort, while costing about a third as much. The company adds that it finished tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol.
Fewer factual errors
OpenAI also reports progress on accuracy. On its internal benchmark, the error rate at low reasoning effort fell from 53.6 per cent for GPT-5.6 Terra to 16.7 per cent for the new model, and dropped to 4.3 per cent at maximum reasoning effort.
Coding gains, with a caveat
On the Terminal-Bench 4.0 coding benchmark, GPT-6.1 Terra solved 50.3 per cent of tasks at maximum reasoning effort. The published score for GPT-5.6 Terra is 21.5 per cent. OpenAI cautioned, however, that GPT-6.1 Sol and GPT-6 Astra remain the better options for the most demanding coding work.
Availability and pricing
GPT-6.1 Terra is open to Plus, Pro, Enterprise and Edu subscribers through ChatGPT Work and Codex. Developers can reach it through the OpenAI API as gpt-6.1-terra. API pricing is $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens. OpenAI says standard API rates are 50 per cent lower than those of GPT-5.6 Terra and GPT-6.1 Sol.
