OpenAI has launched a new AI model, GPT-6.1 Terra, designed not just to answer questions but to work on the computer itself. According to the company, the model understands what is on the screen, takes actions across different apps and writes naturally, and is aimed at everyday work such as filling in forms, working with technical drawings and editing video.
Half the price
The big news is the price. In OpenAI's API, the model will cost $1 per million input tokens and $5 per million output tokens. According to the company, that is half the price of GPT-5.6 Terra and of GPT-6.1 Sol, the more advanced model. It is available now to Plus, Pro, Enterprise and Edu subscribers in ChatGPT Work and Codex, and to developers through the API.
Coding performance
According to OpenAI, on Terminal-Bench 4.0, which tests solving command-line tasks, the model succeeded on 50.3% of tasks at maximum effort, compared with only 21.5% for the previous generation. Even so, the company still recommends GPT-6.1 Sol or GPT-6 Astra for the most complex coding tasks. Keep in mind that all performance and cost figures come from OpenAI's own testing and have not been independently verified.
The race for AI agents
The launch sharpens competition over "computer use", the ability of AI models to operate a computer like a person. Anthropic, Google and Microsoft are developing similar capabilities, each trying to make its agents cheap and reliable enough for businesses to trust them in day-to-day work. The sharp price cut shows the race is no longer just about quality, but also about cost.
What it means for you
For users with a paid ChatGPT subscription, the new model should be available now. For companies and developers, a lower price makes automation tasks that were too expensive worthwhile. As always with AI agents, it is worth starting with simple tasks and checking the results before letting the model act on its own in sensitive systems.
