OpenAI has launched GPT-6.1 Terra, a new language model featuring visual analysis capabilities designed for automating workplace tasks and computer control. The model offers increased speed compared to the previous GPT-5.6 Terra version.
The standard API pricing for GPT-6.1 Terra is set at $1 per million input tokens, $0.05 per million cached input tokens, and $5 per million output tokens. This represents a 50% reduction in cost compared to GPT-5.6 Terra and GPT-6.1 Sol.
In the Agents’ Last Exam benchmark, Terra outperformed GPT-5.6 Terra by 15 percentage points in reasoning while being three times cheaper. At maximum reasoning settings, it matches the accuracy of the flagship GPT-6 Astra but at one-tenth the token cost.
The model shows improved reliability, with error rates in internal tests dropping from 53.6% in GPT-5.6 Terra to 16.7% in the new version at minimal reasoning effort. At maximum effort, the error rate decreased to 4.3%.
Performance in Terminal-Bench 4.0, which tests command-line task execution, more than doubled to 50.3% compared to the 21.5% achieved by its predecessor. The model is now available to ChatGPT Plus, Pro, Enterprise, and Edu subscribers.
