OpenAI announced GPT-6.1 Terra on October 26, presenting it as a model for understanding images and working on computers, with standard API prices 50% below GPT-5.6 Terra’s.
It is designed for practical tasks such as reading screen content, navigating software, filling out forms, understanding technical drawings and drafting text. The model accepts text and images. OpenAI lists API prices of $1 per million input tokens and $5 per million output tokens.
Actual costs depend on usage and applicable fees. OpenAI says Terra scored 15 percentage points higher than its predecessor on the company’s Agents’ Last Exam benchmark when using the medium reasoning setting. The company says Terra did so at about one-third the cost per task.
The company also reports that Terra completed benchmark tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol. These are OpenAI’s results, not an independent assessment. The model is available through the API and in ChatGPT Work and Codex.
AWS has announced that it is available on Amazon Bedrock. The launch presents Terra as a more affordable option for routine work involving screen understanding and actions across applications.