GPT-6.1 Terra is OpenAI's half-price model for computer-use work

OpenAI released GPT-6.1 Terra, the new version of its cheaper, faster model line, which sits below GPT-6.1 Sol and GPT-6 Astra. OpenAI pitches it at the dull, high-volume work of an office day: filling out forms, reading invoices and typing them into accounting software, interpreting technical drawings, drafting text.

The price. In the API, gpt-6.1-terra costs $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens. OpenAI says that is 50% lower than both GPT-5.6 Terra and GPT-6.1 Sol. Two things on the model page to watch: prompts over 272K input tokens are billed at 2x input and 1.5x output for the whole request, and Batch and Flex are half price. The context window is 1.05 million tokens (922K max input, 128K max output), with text and image input only. Tool calling needs the Responses API.

What OpenAI says it can do. All of these are OpenAI's own benchmarks, run by OpenAI:

• Computer use: on Agents' Last Exam at medium reasoning effort, Terra scores 15 percentage points higher than GPT-5.6 Terra at about a third of the cost, and performs "on par with GPT-6 Astra" at a tenth of Astra's price per token.
• Coding: on Terminal-Bench 4.0, a command-line coding test, it solves 50.3% of tasks at max effort, up from GPT-5.6 Terra's 21.5%.
• Fewer made-up answers: on OpenAI's internal set of error-inducing prompts, the error rate at low effort fell from 53.6% to 16.7%. At max effort it is 4.3%.
• Speed: OpenAI says Terra finishes Agents' Last Exam tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol.

Where you get it. It is available now to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and in the API. OpenAI says the same ChatGPT plan allowance covers more tasks with Terra than with Sol or Astra.

The catch.

• OpenAI itself says that "for the most complex coding tasks, GPT-6.1 Sol and GPT-6 Astra remain the better choice." Terra is for volume, not your hardest bugs.
• Its system card addendum says OpenAI is treating Terra as Critical in cybersecurity and High in biological and chemical capability under its Preparedness Framework, so it ships with the same safeguards stack as GPT-6.1 Sol and GPT-6 Astra. That is OpenAI's highest cyber rating on a model built to be cheap enough to run at scale, so builders doing security-adjacent work should expect Astra-level guardrails, not a looser budget model.
• "On par with Astra" is one benchmark, picked and run by the vendor. Independent results will matter more than the launch chart.

Our take. The price cut is the real news. If you are running agents that click through screens or chew through documents all day, a model at $1/$5 that OpenAI says keeps up with its flagship on computer use is worth a test on your own tasks this week. Just don't swap it in for your hardest coding work, because OpenAI doesn't think you should either.

What we did and didn't do: we read OpenAI's announcement, the GPT-6.1 Terra API model page and the system card addendum on October 27, 2026. We have not tested GPT-6.1 Terra ourselves; every benchmark number here is OpenAI's. Not sponsored; no affiliate links.