GPT-6.1 Terra focuses on efficiency: Astra-level at one-tenth price
The post breaks down Terra: 50% lower API prices, better screen understanding and cross-app actions, Astra-level Agents' Last Exam, 3x faster than Astra and Sol; Sol or Astra still recommended for complex coding.
OpenAI just introduced GPT-6.1 Terra, and the biggest thing here is efficiency A few things that stand out: > 50% lower standard API prices compared to GPT-5.6 Terra. > Better at computer use, understanding what’s on screen and taking action across apps. > Substantially better performance on Agent’s Last Exam, which tests how well AI agents handle real work. > According to OpenAI, it can match GPT-6 Astra on this benchmark at one-tenth the price per token. > It also finishes tasks over 3× faster on average than GPT-6 Astra and GPT-6.1 Sol. Terra looks like a practical option for people who want AI agents handling more everyday work without spending as much. For complex coding and in-depth research, OpenAI still recommends Sol or Astra. https://x.com/OpenAI/status/2114780711664398349/video/1