How to Use Gemini 4 Carbon in Factory Droid
Gemini 4 Carbon suits difficult debugging, architectural decisions, and scientific analysis. Compared with Gemini 4 Argon, it spends more reasoning tokens exploring possibilities and checking its work.
Prompts to try with Gemini 4 Carbon in Factory Droid
Copy an example into Factory Droid, or ask in your own words.
Use Gemini 4 Carbon with high reasoning to investigate the failing tests in this repository. Trace the cause across packages, make the smallest justified fix, and run the relevant tests.
Use Gemini 4 Carbon to plan this repository's dependency migration. Map affected packages, order the changes, and define a rollback and verification plan.
Use Gemini 4 Carbon to review the current diff for concurrency bugs. Cite exact code locations and add a regression test for any confirmed defect.
Use Gemini 4 Carbon with high reasoning to compare two ways to remove the circular dependencies in this project. Explain tradeoffs and recommend a staged change.
Use Gemini 4 Carbon to inspect this project's numerical algorithm. Check boundary conditions and units, then test suspected errors before changing code.
How to set up Gemini 4 Carbon in Factory Droid
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the server
Droid saves it to your user config, ~/.factory/mcp.json, so it works in every project.
droid mcp add spiral https://api.spiral.ai/mcp --type http
- Sign in
Start droid, type /mcp, select spiral and authenticate it. Sign in to Spiral in the browser and click Authorize. No API key.
- Check it
Spiral shows as connected. needs authentication means the sign-in did not finish: run /mcp again.
droid mcp list
Use the Spiral skill
- Send Factory Droid 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 Gemini 4 Carbon
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": "vertex", "model": "google/gemini-4-carbon", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Factory Droid finds Gemini 4 Carbon on Spiral, shows the price and runs it.
Gemini 4 Carbon providers and prices
Spiral runs Gemini 4 Carbon through Vertex AI. Open a card for today's price, with every fee included.
Up to $0.106678 per run at the default inputs. $5.20 per 1M input tokens, $0.26 per 1M cached input tokens, $26.00 per 1M output tokens.
Settings you can ask for
Ask Factory Droid for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Max tokens | 1 to 65536 | The most output tokens to generate. | |
| Reasoning effort | low, medium, or high | medium | How much the model reasons. |
Every API field of Gemini 4 Carbon is on its tool page: open its card above.
What can you do with Gemini 4 Carbon in Factory Droid?
Gemini 4 Carbon can investigate bugs across a repository and work through difficult architectural changes. Give it acceptance criteria and ask it to verify the result against the project's tests.
Trace a bug across services
Use Gemini 4 Carbon to connect symptoms across subsystems, compare possible causes, and propose a fix with a regression test.
Plan a migration across packages
Map dependencies, sequence changes, and define checks for a migration that affects several packages.
Review an architectural redesign
Compare designs against stated constraints and identify failure modes before choosing an implementation.
Investigate experimental anomalies
Examine supplied measurements, separate observations from hypotheses, and propose tests that could explain an unexpected result.
Check a mathematical proof
Review a proposed argument, identify unsupported steps, and look for counterexamples or missing assumptions.
Verify a difficult code change
Inspect a patch against its requirements, test edge cases, and report what the available evidence actually verifies.
More on Gemini 4 Carbon and Factory Droid
Gemini 4 Carbon in other agents
Questions about Gemini 4 Carbon in Factory Droid
Can Factory Droid use Gemini 4 Carbon?
Yes. Add Spiral to Factory Droid as an MCP server, and Factory Droid can find and run Gemini 4 Carbon like any built-in tool.
How much does Gemini 4 Carbon cost in Factory Droid?
Up to $0.106678 per run at the default inputs. $5.20 per 1M input tokens, $0.26 per 1M cached input tokens, $5.20 per 1M cache write tokens, $26.00 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Factory Droid plan is separate.
Do I need a Vertex AI account to use Gemini 4 Carbon in Factory Droid?
No. Spiral runs Gemini 4 Carbon on its own Vertex AI account. You sign in to Spiral once and never paste a key.
How long does it take to set up Gemini 4 Carbon in Factory Droid?
About two minutes: add the server, then sign in, then check it.
Does Factory Droid wait while Gemini 4 Carbon runs?
Gemini 4 Carbon runs in the background. Factory Droid starts the run, checks on it and gives you the text, json when it's ready.
What is Gemini 4 Carbon best suited for?
Difficult debugging, architectural redesigns, and work that benefits from repeated verification. Google positions it for demanding tasks beyond the everyday engineering work it recommends Gemini 4 Argon for.
How much context can Gemini 4 Carbon handle?
The context window holds 1,048,576 tokens shared by input, internal reasoning, and output. Leave space for the analysis and answer when supplying a large amount of material.
How long can a Gemini 4 Carbon response be?
Google documents a default maximum of 65,536 output tokens. Internal reasoning also consumes the response budget, so the visible answer can be shorter.
When should I use high reasoning?
Use high reasoning for difficult investigations or proofs where extra checking matters. It gives Gemini 4 Carbon its maximum reasoning depth; ask for a concise final answer separately.
Does Gemini 4 Carbon know current events?
Its knowledge cutoff is March 2026, and some subject knowledge may be older. Supply current evidence for questions about later developments and check factual claims against sources.
Start using Gemini 4 Carbon in Factory Droid
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Factory Droid:
set up https://spiral.ai/SKILL.md


