In the early hours of October 27, 2026, OpenAI announced GPT-6.1 Terra through its official account: a fast, efficient mid-tier model with sharper visual reasoning, more precise computer use, and clearer, more natural writing, built for tasks ranging from decoding technical drawings and drafting proposals to filling in forms and editing videos — the whole workday.
Who gets it first, and what does it cost?
The company says it is rolling out today to Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex. Third-party evaluator Artificial Analysis noted on the same day that this is the first Terra-class model since GPT-5.6, priced at $1 per million input tokens and $5 per million output tokens: input costs half the $2 charged for GPT-5.6 Terra, and output costs less than half its $12.
Why it matters
The Terra name is not new: it surfaced in Microsoft's evaluation configs in mid-October, when it was still just a codename. Its official launch a little over a week later shows the mid-tier lineup is turning over faster. For users, what matters is less the spec sheet than the slot it occupies — not a flagship showpiece, but a tool opened every day for reading diagrams, filling forms, editing video, and drafting proposals, where steadiness and price beat a single record score.
Who should try it now, and who can wait
Paying ChatGPT Work or Codex users will find it where they already work, and it suits everyday copy, forms, and video-editing tasks first. Price-sensitive teams with heavy usage should run the numbers: the same budget now stretches across more tasks. One caveat remains: public information so far centers on the official positioning and third-party price comparisons, with full evaluation details still to come, so test it against your own two or three most common tasks before committing.
