AitopkitOpenAI released GPT-6.1 Terra on Monday, October 26, 2026 — a mid-tier model built for visual reasoning, computer use and clearer writing, aimed squarely at everyday office work. It is available to Plus, Pro, Business, Enterprise and Edu users in ChatGPT Work and Codex, and to developers through the API as gpt-6.1-terra. It is not in regular Chat, and no free tier is listed.
The headline number is the price. Standard API pricing is $1 per million input tokens and $5 per million output tokens (cached input $0.05), which OpenAI says is 50% lower than both GPT-5.6 Terra, the model it replaces, and GPT-6.1 Sol. The context window is 1,050,000 tokens with up to 128,000 output tokens and a June 1, 2026 knowledge cutoff. On OpenAI’s own benchmarks it beats the previous Terra across the board and, on the Agents’ Last Exam computer-use test, matches GPT-6 Astra at roughly a tenth of the price per token — vendor figures that independent tests have not yet verified.
The use cases in the announcement are deliberately unglamorous: reading invoices into accounting software, filling out forms, drafting proposals, reading technical drawings and editing video. That is the point. Terra is not pitched as a research breakthrough; it is pitched as the model you let loose on the repetitive work that eats an afternoon, at a price per task that makes agentic workflows affordable for ordinary businesses rather than just labs.
The limits matter too. Terra accepts text and image input and produces text — it does not take audio or video files directly, and operating video-editing software through its screen is a different capability from understanding video. Prompts above 272,000 input tokens are billed at higher long-context rates. And OpenAI is clear that Sol and Astra remain better for the most complex coding tasks. For high-volume computer work, though, Terra just reset the price floor — and every rival model launch from here will be judged against $1 per task economics.