OpenAI has launched its mid-tier artificial intelligence model GPT-6.1 Terra and begun offering it through ChatGPT Work, Codex and a developer API.

Decrypt, a blockchain media outlet, reported on Oct. 26 that GPT-6.1 Terra is aimed at everyday office tasks such as reading on-screen content, clicking apps and drafting reply outlines.

The model is available to Plus, Pro, Enterprise and Edu users through ChatGPT Work and Codex. Developers can use it via the API. Free users will get a lower-tier model, GPT-5.6 Luna. OpenAI said it halved the standard API fees applied when developers connect the model to their own apps.

The core of the performance improvement is accuracy and processing speed. In an internal evaluation using prompts designed to induce mistakes, OpenAI said the factual error rate fell to 16.7 percent from 53.6 percent for the existing model, GPT-5.6 Terra, when inference strength was set low. Under maximum inference conditions, it recorded 4.3 percent, down 18 percentage points from the previous model.

Its coding-agent performance also improved. In Terminal-Bench 4.0, which assesses the ability to perform tasks in a command-line environment, GPT-6.1 Terra posted a task completion rate of 50.3 percent under maximum inference conditions. That far exceeds the previous model's 21.5 percent. OpenAI said GPT-6.1 Terra also processes tasks on average more than three times faster than GPT-6 Astra and GPT-6.1 Sol. It said those figures are all based on OpenAI's own evaluations, and the company recommends GPT-6.1 Sol or GPT-6 Astra for in-depth research or complex coding work.

OpenAI previously released Terra on a limited basis in June as part of the GPT-5.6 lineup, and formally launched it in July by highlighting price competitiveness. In September, it announced GPT-6 Sol and Luna shortly after Anthropic released Claude Opus 5.5, but did not introduce a new Terra model.

This launch of GPT-6.1 Terra is centred on simultaneously highlighting price cuts and performance improvements. OpenAI clarified Terra's role in everyday work and computer-operation tasks, while maintaining a separation of use cases from Sol and Astra.

GPT-6.1 Terra continues to advance the cost-intelligence frontier across our model families. It’s a significant step up over GPT‑5.6 Terra, with 50 percent lower standard API prices. In ChatGPT Work and Codex, your usage allowance goes further with GPT‑6.1 Terra than with GPT‑5.6 Terra—and far more than with GPT‑6.1 Sol or GPT‑6 Astra. For in-depth research or complex coding, choose GPT-6.1 Sol or GPT-6 Astra.