OpenAI releases GPT-6.1 Terra, a cheaper model built to handle routine computer work
The model can read a screen and click, type and switch between apps, The Next Web reported. Its performance figures come from OpenAI's own testing.
OpenAI has released GPT-6.1 Terra, a faster and cheaper AI model aimed at the routine computer work of a typical office day, according to The Next Web. This story is based on that one report.
The Next Web reported that the new model replaces GPT-5.6 Terra. It sits in the middle of OpenAI's range, below Sol and Astra but above Luna.
The main change is what OpenAI calls computer use. Terra can look at a screen, understand what it shows, and act on it by clicking, typing and moving between apps. In one example, a company could give it a stream of incoming invoices, and Terra would read each one and type the details into accounting software. OpenAI's other examples include filling in forms, drafting proposals, reading technical drawings and editing videos.
Terra is available to ChatGPT Plus, Pro, Enterprise and Edu users and in Codex, OpenAI's coding tool. Developers pay $1 per million input tokens and $5 per million output tokens, half the price of GPT-5.6 Terra.
The launch matters mainly because of what OpenAI says the model can do at that price. In one test, Terra watched an animated film shot by shot and wrote a time-coded log for a restoration team. OpenAI said it scored 15 percentage points higher than its predecessor at about a third of the cost. The company said Terra also matched GPT-6 Astra, its most capable model, at a tenth of the price.
OpenAI also said Terra makes fewer factual mistakes. On a set of questions designed to trip models up, its error rate at the lowest setting fell from 53.6% to 16.7%. In a test of coding at the command line, it solved about half the tasks, more than double its predecessor's score. OpenAI still recommends Sol or Astra for the hardest coding work.
On safety, OpenAI said Terra reaches the "Critical" level for cybersecurity skills under its Preparedness Framework, the system it uses to rate risky abilities. It carries the same protections as GPT-6 Astra. The release follows a setback last month, when OpenAI cancelled GPT-6.1 Astra after it failed internal safety tests. OpenAI said Terra is more likely than its predecessor to stay within the task it is given and to tell users about its limits.
Several questions remain open. The performance figures come from OpenAI. The company also compared Terra with Anthropic's Claude Opus 5.5 and said Terra completed tasks at roughly a ninth of the cost. The Next Web noted that those results come from OpenAI's own testing and have not been independently checked.
The report does not name most of the tests used. It does not describe how the cybersecurity protections work, and it does not say what went wrong in the GPT-6.1 Astra safety tests.
