OpenAI launches GPT-6.1 Terra for cheaper computer use

OpenAI released GPT-6.1 Terra, a fast model built for computer use tasks like filling out forms, reading technical drawings, and editing videos, alongside clearer writing. The company says standard API prices are 50% lower than its predecessor, GPT-5.6 Terra, and than GPT-6.1 Sol, which OpenAI says makes high-volume automation, such as reading invoices into accounting software, more practical.

On OpenAI's own benchmarks, the model scores higher than GPT-5.6 Terra on a professional-tasks benchmark called Agents' Last Exam while costing a fraction as much, and the company says it is far more token-efficient than Claude Opus 5.5 at top reasoning settings. OpenAI also reports a sharp drop in factuality errors on its internal error-inducing prompt set, from 53.6% to 16.7% at low reasoning effort. On a coding benchmark called Terminal-Bench 4.0, OpenAI says the model solves more than double the tasks GPT-5.6 Terra could, though it says GPT-6.1 Sol and GPT-6 Astra remain better for the hardest coding work.

OpenAI says the model meets its Critical threshold for cybersecurity risk under its Preparedness Framework and carries safety protections built for GPT-6 Astra. All of these benchmark figures are OpenAI's own. GPT-6.1 Terra is available starting today to ChatGPT Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, and via the API as gpt-6.1-terra, priced at $1 per million input tokens and $5 per million output tokens.

Why it matters

Cheaper, faster computer-use automation could make agentic workflows like document processing far more affordable to run at scale.

Source: OpenAI