OpenAI releases GPT-6.1 Terra, its cheaper mid-tier model

OpenAI released GPT-6.1 Terra, a mid-tier model that it described as faster and more efficient than its flagships and as “a significant step up” from GPT-5.6 Terra. Standard API prices are $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens, which OpenAI said is 50% lower than both GPT-5.6 Terra and GPT-6.1 Sol. It is available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and through the API as gpt-6.1-terra.

Every performance figure in the announcement is OpenAI’s own, and the post does not cite independent testing. On Agents’ Last Exam, a benchmark of professional computer-use tasks, OpenAI said Terra at medium reasoning effort scores 15 percentage points above GPT-5.6 Terra at about a third of the cost, and performs on par with GPT-6 Astra at a tenth of its per-token price. On Terminal-Bench 4.0, a command-line coding test, it reported 50.3% of tasks solved at maximum effort, against a “published score” of 21.5% for GPT-5.6 Terra. It also said Terra is more than three times as token-efficient as Claude Opus 5.5 at a quarter of the per-token price, “resulting in a ~9x lower cost per task”. For the most complex coding work, the company said, Sol and Astra remain the better choice.

OpenAI also pointed to fewer factual errors. On an internal benchmark of error-inducing prompts, the error rate at low reasoning effort fell from 53.6% for GPT-5.6 Terra to 16.7%, and at maximum effort fell by 18 percentage points to 4.3%. The test is OpenAI’s own and its contents are not published.

The safety section carries the least expected line. OpenAI said Terra “meets the Critical threshold in cybersecurity under our Preparedness Framework” and keeps the protections developed for Astra, the first model it had rated Critical for cyber capability. That puts a model priced for high-volume work in the highest tier of OpenAI’s cyber classification, with the safeguards applied to the larger model carried over. OpenAI said its evaluations show Terra is more likely than GPT-5.6 Terra to stay within scope and to disclose limitations to the user, and referred readers to a system card for detail.

Terra fills the middle of the GPT-6 line that OpenAI assembled over the preceding weeks, following the GPT-6 Sol and Luna launch and the DevDay release of GPT-6.1 Sol. The comparisons with Opus 5.5 are OpenAI’s; the sources consulted contained no independent test of them.