How to Use GPT-6.1 Terra in Claude Code
GPT-6.1 Terra fits screenshot analysis, concise summaries, and narrowly scoped bug fixes. Choose it for everyday work; Terra Pro uses the same underlying model with additional reasoning for harder tasks.
Prompts to try with GPT-6.1 Terra in Claude Code
Copy an example into Claude Code, or ask in your own words.
Use GPT-6.1 Terra with medium reasoning to reproduce the failing test in this project, find its cause, and make the smallest fix. Run the relevant regression test and summarize the result.
Use GPT-6.1 Terra to compare the current page screenshot with the reference image in this project. Fix the three clearest layout discrepancies and verify them in a fresh screenshot.
Use GPT-6.1 Terra to turn this project's recent commit notes into a release summary under 180 words. Include user-visible fixes and preserve any stated limitations.
Use GPT-6.1 Terra with medium reasoning to inspect the settings screen and its component code. Fix the label mismatch, verify the displayed result, and report which files changed.
Ask GPT-6.1 Terra through Spiral to review the diff on this branch and list the three riskiest changes. Save the result in ./assets.
How to set up GPT-6.1 Terra in Claude Code
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the hosted server
This adds Spiral to the current project. Add --scope user to use it in every project.
claude mcp add --transport http spiral https://api.spiral.ai/mcp
- Sign in
Start claude, type /mcp, choose spiral and Authenticate. Sign in to Spiral in the browser and click Authorize. No API key.
Use the Spiral skill
- Send Claude Code this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run GPT-6.1 Terra
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openai", "model": "openai/gpt-6.1-terra", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Claude Code finds GPT-6.1 Terra on Spiral, shows the price and runs it.
GPT-6.1 Terra providers and prices
Spiral runs GPT-6.1 Terra through OpenAI. Open a card for today's price, with every fee included.
Up to $0.026682 per run at the default inputs. $1.30 per 1M input tokens, $0.065 per 1M cached input tokens, $1.625 per 1M cache write tokens, $6.50 per 1M output tokens.
Settings you can ask for
Ask Claude Code for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Top p | a number up to 1 | Taken only with reasoning effort "none". | |
| Logprobs | on or off | Taken only with reasoning effort "none". | |
| Max tokens | 16 to 128000 | The most output tokens to generate. | |
| Temperature | 0 to 2 | Taken only with reasoning effort "none". | |
| Top logprobs | 0 to 20 | Taken only with reasoning effort "none". | |
| Presence penalty | -2 to 2 | Taken only with reasoning effort "none". | |
| Reasoning effort | none, low, medium, high, xhigh, or max | medium | How much the model reasons. |
| Frequency penalty | -2 to 2 | Taken only with reasoning effort "none". | |
| Parallel tool calls | on or off |
Every API field of GPT-6.1 Terra is on its tool page: open its card above.
What can you do with GPT-6.1 Terra in Claude Code?
Turn a reproducible bug or a screenshot discrepancy into a focused code fix. Give the agent the relevant project files, the expected behavior, and a check that proves the change works.
Screenshot review
Ask Terra to explain a settings screen, identify visible discrepancies, or summarize a dashboard. Supply enlarged text where labels or numbers are hard to read.
Focused bug fixes
Give Terra a reproducible failure and a clear expected result. Ask for a limited patch and a regression check that demonstrates the fix.
Concise summaries
Turn supplied notes into a short decision brief, handoff, or customer explanation. Specify the audience, length, and details the summary must preserve.
Review of a difficult decision
Try Terra Pro on a bounded problem with conflicting evidence or several competing explanations. Request a recommendation tied to the supplied evidence and an explicit list of unresolved questions.
Desktop record checks
Use Terra to inspect visible records against stated criteria. Define the expected final state and ask the agent to verify it before reporting completion.
Comparison of proposed fixes
Ask Terra Pro to compare two fixes against the same reproduction case, constraints, and regression requirements. Request the smallest change that satisfies those requirements.
More on GPT-6.1 Terra and Claude Code
Overview
GPT-6.1 Terra in other agents
More Claude Code guides
Questions about GPT-6.1 Terra in Claude Code
Can Claude Code use GPT-6.1 Terra?
Yes. Add Spiral to Claude Code as an MCP server, and Claude Code can find and run GPT-6.1 Terra like any built-in tool.
How much does GPT-6.1 Terra cost in Claude Code?
Up to $0.026682 per run at the default inputs. $1.30 per 1M input tokens, $0.065 per 1M cached input tokens, $1.625 per 1M cache write tokens, $6.50 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Claude Code plan is separate.
Do I need an OpenAI account to use GPT-6.1 Terra in Claude Code?
No. Spiral runs GPT-6.1 Terra on its own OpenAI account. You sign in to Spiral once and never paste a key.
How long does it take to set up GPT-6.1 Terra in Claude Code?
About two minutes: add the hosted server, then sign in.
Does Claude Code wait while GPT-6.1 Terra runs?
GPT-6.1 Terra runs in the background. Claude Code starts the run, checks on it and gives you the text, json when it's ready.
What is GPT-6.1 Terra good at?
Use it to interpret screenshots, navigate desktop apps, write concise summaries, and fix narrowly scoped bugs. These are the tasks OpenAI specifically recommends for Terra.
How does Terra Pro differ from Terra?
Terra Pro uses the same underlying model with pro reasoning enabled. Its additional reasoning is intended for complex tasks where answer quality needs extra attention.
How much material can Terra handle, and how long can its answers be?
OpenAI lists a 1,050,000-token context window and a maximum output of 128,000 tokens. Internal reasoning shares the generated-token budget, so that maximum is not a promise of 128,000 tokens of visible text.
Can Terra create images, audio, or video directly?
Terra directly produces text and accepts text and images. Its model specification does not list direct audio or video support.
What should I do when a screenshot contains tiny text?
Enlarge the relevant text before asking Terra to read it. Ask it to identify anything unreadable so your task does not depend on a guessed label or number.
Start using GPT-6.1 Terra in Claude Code
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Claude Code:
set up https://spiral.ai/SKILL.md


