News
Breaking: OpenAI Launches GPT-6.1 Terra, Halving API Prices for Computer-Use AI
OpenAI released GPT-6.1 Terra on October 26, 2026: a mid-tier model for computer use at $1/$5 per million tokens, half the price of its predecessor.
OpenAI released GPT-6.1 Terra on Monday, October 26, 2026, a new mid-tier AI model built for computer use: reading screens, filling in forms, entering data, and handling other routine office work. The headline is price. Through OpenAI's API, Terra costs $1 per million input tokens and $5 per million output tokens, with cached input at $0.05 per million — about half the price of the previous GPT-5.6 Terra and of the larger GPT-6.1 Sol model.
What we know
- Terra is available now to Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, and to developers through the API as gpt-6.1-terra. Amazon Web Services made it available on Bedrock the same day.
- It is not in regular ChatGPT, and OpenAI's announcement names paid plans only — there is no free tier listed for Terra.
- OpenAI positions Terra for high-volume screen-based work: drafting proposals, reading technical drawings, filling out forms, and editing video, rather than the hardest coding tasks, where it still points users to GPT-6.1 Sol or GPT-6 Astra.
- On OpenAI's own computer-use benchmark, Agents' Last Exam, Terra scored 15 percentage points higher than GPT-5.6 Terra at medium reasoning effort while costing roughly a third as much per task, and OpenAI says it finishes tasks more than three times faster than Astra or Sol.
- On Terminal-Bench 4.0, a coding test, OpenAI reports Terra solving 50.3% of tasks at maximum effort, up from 21.5% for GPT-5.6 Terra. On its internal error-provoking prompt set, factual errors fell from 53.6% to 16.7% at low reasoning effort, and to 4.3% at maximum effort.
- Independent testing firm Artificial Analysis scored Terra at 48 on its Intelligence Index at maximum effort, up from 42 for GPT-5.6 Terra, at $0.40 per task versus $1.40 before.
What's unconfirmed
All benchmark and error-rate figures above come from OpenAI's own measurements or from early third-party reports of them; they have not yet been independently verified across real-world workloads. How much of the claimed speed and cost advantage holds up in everyday business use — as opposed to benchmark conditions — is still an open question, and this flash will not overstate it.
Why it matters
For developers and businesses, Terra cuts the cost of automating screen-based work roughly in half versus its predecessor, which could bring computer-use agents within reach of smaller teams. For regular ChatGPT users, nothing changes today: Terra is not in standard Chat, and ChatGPT's plan prices are untouched by this launch. The bigger signal is the direction of travel — capable AI models keep getting cheaper, and the mid-tier model, not the flagship, is where most of the everyday volume will likely run.
Sources: OpenAI's GPT-6.1 Terra launch announcement, as reported by BleepingComputer, KuCoin News, BeInCrypto, and Artificial Analysis data cited in launch coverage. Benchmark figures are OpenAI's own unless noted. This is not financial advice.