OpenAI launches GPT-6.1 Terra with better visual understanding to fill forms or edit video
OpenAI unveiled GPT-6.1 Terra, a new version of its models that improves visual comprehension for interacting with a computer, allowing it to complete forms, analyze technical blueprints and edit video. According to the company, the model gained 15 points over its previous version on the Agents Last Exam benchmark and cut its error rate at low reasoning from 53.6% to 16.7%. Its cost per token is about one-tenth that of OpenAI's most advanced model.
In 35 seconds
- GPT-6.1 Terra strengthens visual comprehension for tasks such as completing forms, interpreting technical blueprints and editing video.
- OpenAI reports a 15-point gain on the Agents Last Exam benchmark and performance equivalent to GPT-6 Astra.
- The error rate at low reasoning fell from 53.6% to 16.7% in an internal test.
IASummary generated by artificial intelligence from a story by Revista Semana. It may contain errors: the original source prevails.




Comments
Be respectful. Comments reported by several people are hidden until a moderator reviews them.
You do not need an account to read, only to comment. Sign in with a code sent to your email or with your Google account. We only show the name you choose, never your email.