OpenAI Update: GPT-6.1 Terra with Eternal Discount

On October 26, OpenAI released GPT-6.1 Terra, an update to the mid-range model of the lineup, which, according to the company, costs developers 50% less than GPT-5.6 Terra and handles computing tasks faster and more accurately. Terra is an everyday model for routine office work: reading what's on the screen, navigating applications, and composing responses. This skill is called computer usage: the AI sees the screen and clicks and types just like a user.

"This is a significant step forward compared to GPT-5.6 Terra, with standard API prices 50% lower," OpenAI stated. "GPT-6.1 Terra is already being deployed and is available to all Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex." Free users receive GPT-5.6 Luna.

In OpenAI's internal test with error-inducing queries, the rate of actual errors with low reasoning effort decreased from 53.6% for GPT-5.6 Terra to 16.7%. On Terminal-Bench 4.0, Terra solved 50.3% of tasks with maximum effort compared to 21.5% for its predecessor. However, it lags behind GPT-6.1 Sol (55.9%), GPT-6 Astra (58.2%), and Claude Sonnet 5.5 (70.6%), and OpenAI itself recommends Sol or Astra for the most complex coding tasks. On MMMU-Pro (Vision), Terra scores 85.5% compared to 86.3% for Sol and 86.7% for Astra. All figures are OpenAI's own data, with the error test being internal.

The standard API price is $1 per million input tokens, $0.05 for cached input, and 5 per million outputs. Batch and Flex are 50% cheaper, Fastmode is twice as expensive. This is half the price of GPT-6.1 Sol (2/10) and one-tenth the price of GPT-6 Astra (10/$50). The context is 1.05 million tokens, output is up to 128K, with a knowledge cutoff of June 1, 2026.