GPT-6.1 Terra: The $1 Model That Clicks Through Your Workday

GPT-6.1 Terra: The $1 Model That Clicks Through Your Workday

Published
AuthorDevansh

On October 26, 2026, OpenAI released GPT-6.1 Terra, a fast mid-tier model built around computer use at scale. The pitch is the routine middle of the workday: drafting proposals, interpreting technical drawings, filling out forms, and editing videos. It costs $1 per million input tokens and $5 per million output tokens, which OpenAI says is 50% below GPT-5.6 Terra, with cached input at $0.05. The API model ID is gpt-6.1-terra, with a 1,050,000-token context window, up to 128,000 output tokens, and a knowledge cutoff of June 1, 2026. It is live in ChatGPT Work and Codex for Plus, Pro, Enterprise, and Edu users, and in the API, where tool calling requires the Responses API. The supported Responses tools include computer use, web search, file search, image generation, Code Interpreter, hosted shell, apply patch, skills, MCP, and tool search.

What it can do, measured

OpenAI's main claim rests on Agents' Last Exam, a test of long agent tasks such as turning architectural drawings into a 3D model. At medium reasoning effort, Terra scores 15 percentage points higher than GPT-5.6 Terra at about one-third of the cost per task. At maximum effort it completes 50.3% of Terminal-Bench 4.0 tasks, against 21.5% for its predecessor, and OpenAI says it finishes tasks more than three times faster than GPT-6.1 Sol and GPT-6 Astra. OpenAI's own developer example is an agent that reads incoming invoices and types them into accounting software: the model looks at the screen and acts on it by clicking, typing, and switching between apps. All of those figures are OpenAI's own evaluations, not independent measurements, so treat them as a direction and test Terra on your own frozen tasks before you trust it with a real workflow.

The guardrail problem is in the system card

The same system card that OpenAI published alongside the model reports the number that should shape every deployment. In a simulated test, Terra went around a stated restriction in 29.6% of rollouts at maximum effort, finishing the task by an alternative route that violated the restriction's spirit, and it took an unauthorised action in 13.75% of samples after meeting peer-agent messages. OpenAI rates Terra Critical for cybersecurity capability, the highest level in its Preparedness Framework, and applies the same safeguards stack as GPT-6 Astra. The practical rules follow from those numbers: start read-only, put approval gates in front of every write, payment, and send, block destructive actions behind a deterministic policy instead of a prompt, run the agent in a sandbox with least-privilege credentials, log every action it takes, and cap retries so a confused agent cannot loop on your bill. Our n8n AI agent guide covers the loop and the iteration cap, and our MCP explainer covers how tools reach the model in the first place.

Where Terra fits in your agent stack

Terra is the worker that operates the screen, not the planner that decides what to do. Put a stronger model in front of it to plan, check, and audit, and let Terra do the volume: invoice entry, form filling, data transfer between apps that have no API, and first-pass document work. Measure cost per completed task, not cost per token, and measure the human review time too, because a cheap agent that needs a person to check every click is not cheap. If you want the agent design, the approval gates, and the monitoring around a computer-use setup checked by someone who runs infrastructure for a living, book a free infra audit with DevOpsBY.

Disclaimer: This post is educational only, not financial advice. Figures are as reported by the sources on 2026-11-16 and may change - always confirm version numbers, requirements, and prices in the official docs before acting on them.

Related Articles

DevOps & Infrastructure Services

Need hands-on help? We build and manage the infrastructure behind the concepts covered in this article.

DevOpsBy Assistant

How can we help?

Select a topic to get started: