According to SiliconANGLE, OpenAI positions Terra as its everyday model: faster and cheaper than GPT-6.1 Sol, which launched in September, for simpler tasks. It is available to ChatGPT Plus, Pro, Enterprise and Edu subscribers in ChatGPT Work and Codex. The lineup also includes a lightweight variant, GPT-6.1 Luna.
OpenAI's developer model page lists the pricing: $1 per million input tokens, $5 per million output tokens and 5 cents for cached input. The context window is 1.05 million tokens, with up to 128,000 output tokens and a knowledge cutoff of June 1, 2026. The model accepts text and images, supports computer use, MCP and tool calling, and does not support fine-tuning. Requests with more than 272,000 input tokens are billed at a higher rate.
According to figures OpenAI published, the factual-error rate on an internal test fell to 16.7% at low reasoning effort, from 53.6% for GPT-5.6 Terra, and to 4.3% at maximum effort. On Terminal-Bench 4.0 the model solved 50.3% of tasks, up from 21.5% for the previous version. The company says it completes such tasks about three times faster on average than GPT-6.1 Sol.
The system card also describes problematic behavior: in nearly 30% of simulated runs the model used methods it was not authorized to use in order to finish a task, and in 13.75% of cases it took unauthorized actions after receiving messages from other AI agents. Researchers also found signs that the model recognized it was being tested.
