OpenAI has released GPT-6.1 Terra, a new artificial intelligence model built for a job that goes beyond answering questions: operating a computer. The model is available in ChatGPT Work and Codex for Plus, Pro, Enterprise and Edu users, and developers can call it through the company's API under the identifier gpt-6.1-terra.
Terra is designed for the repetitive, screen-bound work that fills an office day: drafting proposals, interpreting technical drawings, filling out forms and editing video. It combines visual understanding and reasoning with the ability to carry out computer tasks, processing large workloads quickly and at a lower cost than the models above it in OpenAI's lineup.
Half the price, sharper on computer tasks
The headline change is price. Standard API rates are $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens. That is 50 percent lower than the previous model, GPT-5.6 Terra, and the next model up, GPT-6.1 Sol, and roughly one-tenth the token price of the flagship GPT-6 Astra.
Performance has moved in the other direction, according to OpenAI's own evaluations. On Agents' Last Exam, a benchmark that tests professional task execution such as turning architectural drawings into three-dimensional models or analysing video by scene, Terra scored 15 percentage points higher than GPT-5.6 Terra at medium reasoning level, while the cost of completing a task fell to about one-third of its predecessor's.
On Terminal-Bench 4.0, which measures how well coding agents solve problems in a command-line environment, Terra solved 50.3 percent of tasks at maximum reasoning level, more than double GPT-5.6 Terra's published score of 21.5 percent. Those figures are the company's evaluations rather than independent tests, but the direction is clear: cheaper and more capable at once.
The computer-use era, in Altman's words
OpenAI chief executive Sam Altman announced the launch on Tuesday, calling Terra a capable new model and singling out one ambition above all others. In a post on X, he wrote that the company is especially excited for what he called a new era of computer use, a phrase that signals where OpenAI's sharpest commercial bet now sits.
Computer use is the industry's shorthand for agentic AI: systems that navigate graphical interfaces, click buttons, fill forms, switch between applications and complete multi-step tasks without a human doing the legwork. It is a significant leap beyond answering questions or generating text. For everyday users and enterprise software teams alike, it means an AI that does not just advise, but acts.
A tier for every job, and a price war underneath
Terra is the latest step in OpenAI's strategy of giving businesses a menu of models to combine: a high-performance model for complex work, a decision-support model, and now a low-cost specialist for repetitive computer tasks. The company has been widening that menu quickly, including the Decisions API launched on October 22, powered by GPT-6 Luna, which helps agents make straightforward calls such as routing customer inquiries to the right department.
The timing also fits a broader shift in the AI market. With companies questioning the vast sums being poured into AI infrastructure, a model that halves the cost of each task is an easier sell than a bigger, pricier one. Terra's pitch is not that it is the smartest model OpenAI has ever built. It is that it is cheap enough, and capable enough, to be left running on the everyday work that used to need a person at the keyboard.



