How to Use GPT-6.1 Terra for Development

GPT-6.1 Terra, released October 26, 2026, is OpenAI’s fast, lower-cost model for coding, computer use, and everyday work. It is available in Codex and through the API as gpt-6.1-terra.

Great for

Everyday coding, tests, small features, bug fixes, and app testing

Use a larger model for

Complex architecture and the hardest debugging

• Choose Terra as the model for routine tasks
• Keep project instructions in AGENTS.md
• Give clear tasks with tests to prove the result

See our How to Prompt OpenAI Codex guide for prompt patterns.

• Small features and UI changes
• Bug fixes with a failing test first
• Writing and updating tests
• Refactors with unchanged behavior
• Docs and comments

• Have Terra click through flows in your app
• Check forms, sign-up, and checkout
• Ask for screenshots as proof
• Run it in a test environment, not on production data

• Send several small, independent tasks at once
• Keep each task to one change
• Avoid tasks that touch the same files
• Review each result separately

Review this pull request for bugs, security issues, missing tests, and unclear naming. List issues by severity with file and line references.

• Use gpt-6.1-terra for high-volume features such as summaries, extraction, and classification
• Cache repeated instructions to cut costs
• Set limits on output length
• Log usage per feature

Input

$1 per million tokens

Cached input

$0.05 per million tokens

Output

$5 per million tokens

Check OpenAI’s pricing page for current rates before budgeting.

• Keep secrets out of prompts and repositories
• Review every diff before merging
• Limit what computer-use sessions can access
• OpenAI rates Terra at its Critical level for cybersecurity, so follow its usage policies

Plan

Break work into small tasks

Build

Terra handles routine tasks in Codex

Test

Run tests and computer-use checks

Review

You read every diff