OpenAI launches GPT-6.1 Terra, a cheaper AI that can use a computer

OpenAI has released GPT-6.1 Terra, a new AI model it says is fast, efficient and able to use a computer. The company describes the goal as "fast and capable computer use at scale." It also promises clearer, more natural writing than before.

What it does

OpenAI says Terra combines detailed visual reasoning with fast, efficient performance. In plain terms, visual reasoning means the AI can look at what is on a screen and make sense of it. Computer use means it can then act on what it sees, the way a person would with a mouse and keyboard.

What it costs

Pricing is $1 per million input tokens and $5 per million output tokens. Tokens are the small chunks of text an AI reads and writes, and they are how many AI services bill their customers. Input is what you send to the model. Output is what it sends back.

Who it affects

The most direct audience is businesses that pay for AI tools or build products on top of OpenAI's models. A cheaper model that can handle screen-based tasks could lower those bills. Everyday users may see it show up in tools they already use, though the material does not say so.

What is not known yet

The announcement as reported does not say how Terra compares with OpenAI's other models on accuracy or speed. It also does not say where the model is available, when it rolls out, or how it will appear in products like ChatGPT. Those details, and any independent testing, will show whether the claims hold up.