How to Use Claude Opus 4.7 in Google Antigravity
Claude Opus 4.7 fits complex coding, debugging, and work that takes several steps to finish. Compared with Claude Opus 4.6, it improves advanced software engineering and follows instructions more literally.
Prompts to try with Claude Opus 4.7 in Google Antigravity
Copy an example into Google Antigravity, or ask in your own words.
Use Claude Opus 4.7 with extra-high reasoning effort to investigate the failing tests in this repository. Trace the cause, implement a focused fix, and run the relevant checks; report the changed files and results.
Use Claude Opus 4.7 to review the current branch against main. Focus on correctness, compatibility, and failure handling; give each finding a file location, a concrete failure scenario, and a suggested fix.
Use Claude Opus 4.7 with extra-high reasoning effort to plan a migration of this project's persistence layer. Read the existing implementation, preserve public behavior, and define stages with acceptance criteria before editing.
Use Claude Opus 4.7 to trace how this project handles retries. Identify paths that could repeat a completed operation, propose a correction, and add a regression check for the demonstrated failure.
Use Claude Opus 4.7 to compare the implementation with the project's specification. List unmet or contradictory requirements with file references, then propose an ordered repair plan.
How to set up Claude Opus 4.7 in Google Antigravity
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Open the MCP config
* Antigravity App: click Customizations, then Installed. In the MCP Servers section, click Copy Config File Path and open that file. * Antigravity IDE: click … at the top of the agent side panel, select MCP Servers, then Manage MCP Servers › View raw config. * Antigravity CLI: edit ~/.gemini/config/mcp_config.json.
- Add the server
Add Spiral under mcpServers:
{ "mcpServers": { "spiral": { "serverUrl": "https://api.spiral.ai/mcp" } } } - Sign in
In the Antigravity App, open Customizations › Installed. Under MCP Servers, click Authenticate next to Spiral. Sign in to Spiral, pick the organization and click Authorize. Copy the authorization code the browser shows, paste it into Enter authorization code and click Submit. No API key.
Use the Spiral skill
- Send Google Antigravity 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 Claude Opus 4.7
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": "anthropic", "model": "anthropic/claude-opus-4.7", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Google Antigravity finds Claude Opus 4.7 on Spiral, shows the price and runs it.
Claude Opus 4.7 providers and prices
Spiral runs Claude Opus 4.7 through Anthropic. Open a card for today's price, with every fee included.
Up to $0.133575 per run at the default inputs. $6.50 per 1M input tokens, $0.65 per 1M cached input tokens, $8.125 per 1M cache write tokens, $13.00 per 1M 1-hour cache write tokens, $32.50 per 1M output tokens; for a prompt above 200,000 tokens, $13.00 per 1M input tokens, $1.30 per 1M cached input tokens, $16.25 per 1M cache write tokens, $26.00 per 1M 1-hour cache write tokens, $48.75 per 1M output tokens.
Settings you can ask for
Ask Google Antigravity for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Max tokens | 1 to 128000 | The most output tokens to generate. On a reasoning model the reasoning tokens count toward it: a cap too low for the effort can end with no answer (finish_reason length). | |
| Reasoning effort | none, minimal, low, medium, high, xhigh, or max | none | How much the model reasons. |
| Parallel tool calls | on or off |
Every API field of Claude Opus 4.7 is on its tool page: open its card above.
What can you do with Claude Opus 4.7 in Google Antigravity?
Claude Opus 4.7 can carry a difficult code change from investigation through implementation and review. Give it the relevant files, constraints, and acceptance criteria up front.
Trace a bug across services
Ask Claude Opus 4.7 to follow a failing request through the relevant code, identify the cause, and propose a fix with a regression check.
Review a substantial code change
Have Claude Opus 4.7 examine a change across files and report concrete failure scenarios, affected locations, and suggested corrections.
Plan a legacy code migration
Give Claude Opus 4.7 the current design and compatibility requirements, then request migration stages with acceptance criteria for each stage.
Compare technical proposals
Ask Claude Opus 4.7 to compare supplied proposals against explicit requirements and produce a decision table with unresolved questions.
Turn an incident timeline into a review
Provide Claude Opus 4.7 with the incident narrative and diagnostic evidence, then request a causal analysis and proposed follow-up checks.
Check a specification for contradictions
Use Claude Opus 4.7 to compare requirements across a long specification and flag conflicting statements before implementation.
Draft a technical handoff
Ask Claude Opus 4.7 to turn a completed investigation into a concise handoff covering the problem, findings, remaining questions, and next actions.
More on Claude Opus 4.7 and Google Antigravity
Claude Opus 4.7 in other agents
More Google Antigravity guides
Questions about Claude Opus 4.7 in Google Antigravity
Can Google Antigravity use Claude Opus 4.7?
Yes. Add Spiral to Google Antigravity as an MCP server, and Google Antigravity can find and run Claude Opus 4.7 like any built-in tool.
How much does Claude Opus 4.7 cost in Google Antigravity?
Up to $0.133575 per run at the default inputs. $6.50 per 1M input tokens, $0.65 per 1M cached input tokens, $8.125 per 1M cache write tokens, $13.00 per 1M 1-hour cache write tokens, $32.50 per 1M output tokens; for a prompt above 200,000 tokens, $13.00 per 1M input tokens, $1.30 per 1M cached input tokens, $16.25 per 1M cache write tokens, $26.00 per 1M 1-hour cache write tokens, $48.75 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Google Antigravity plan is separate.
Do I need an Anthropic account to use Claude Opus 4.7 in Google Antigravity?
No. Spiral runs Claude Opus 4.7 on its own Anthropic account. You sign in to Spiral once and never paste a key.
How long does it take to set up Claude Opus 4.7 in Google Antigravity?
About two minutes: open the MCP config, then add the server, then sign in.
Does Google Antigravity wait while Claude Opus 4.7 runs?
Claude Opus 4.7 runs in the background. Google Antigravity starts the run, checks on it and gives you the text, json when it's ready.
What is Claude Opus 4.7 best suited for?
Complex coding, debugging, and reviews that require following a problem across several files. Anthropic also recommends it for multistep work that previously needed close supervision.
How much context and output can Claude Opus 4.7 handle?
It has a one-million-token context window and supports up to 128,000 output tokens per standard response. Those are token limits, not word counts.
Should I always choose maximum reasoning effort?
Start with extra-high effort for coding and agent work. Maximum effort can overthink structured-output tasks, so use it when testing shows a useful improvement.
What should I include in a coding prompt?
State the goal, constraints, acceptance criteria, and relevant file locations in the first request. Give the model enough context to complete the task without piecing together requirements across many turns.
How should I organize a long document prompt?
Put the source material first and the question afterward. Include an example when the answer needs a particular format or voice.
Start using Claude Opus 4.7 in Google Antigravity
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Google Antigravity:
set up https://spiral.ai/SKILL.md


