[Integrations](https://spiral.ai/integrations) › [Amp](https://spiral.ai/integrations/amp) › [Gemini 4 Argon](https://spiral.ai/integrations/amp/gemini-4-argon)

# How to Use Gemini 4 Argon in Amp

Gemini 4 Argon suits complex coding, technical investigations, and legal or financial analysis. Compared with Gemini 4 Carbon, it spends fewer thinking tokens per turn.

|  |  |
| --- | --- |
| Model | Gemini 4 Argon |
| Available through | Vertex AI |
| Category | LLM |
| Price | Up to $0.106678 per run at the default inputs. $5.20 per 1M input tokens, $0.26 per 1M cached input tokens, $0.4875 per 1M cache write tokens, $26.00 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Gemini 4 Argon in Amp

- Use Gemini 4 Argon with high reasoning to trace the failing test through this repository. Read the project conventions, identify the root cause, and implement the smallest fix with a regression test.
- Use Gemini 4 Argon to investigate slow startup in this project. Measure a baseline, trace expensive imports, and propose changes with before-and-after measurements.
- Use Gemini 4 Argon with high reasoning to plan a refactor of the session storage layer. Map its callers, preserve public behavior, and specify tests for each migration step.
- Use Gemini 4 Argon to review the current diff for concurrency bugs. Show the execution sequence behind each finding and add a reproducing test where possible.
- Use Gemini 4 Argon to compare this implementation with the project's design documentation. Cite the relevant files and identify behavioral differences, missing tests, and unresolved assumptions.

## How to set up Gemini 4 Argon in Amp

Pick the way that suits you. Each takes about two minutes.

### Use Spiral's MCP server

1. **Add the server**: No header and no client ID or secret: Amp registers itself with Spiral.

```bash
amp mcp add spiral https://api.spiral.ai/mcp
```

2. **Sign in**: Start `amp`. Amp opens Spiral's sign-in page in your browser: sign in, pick the organization and click Authorize. No API key.

3. **Check it**

```bash
amp mcp doctor
```

### Use the Spiral skill

1. **Send Amp this message**: Your agent reads Spiral's skill and sets itself up.

```
set up https://spiral.ai/SKILL.md
```

2. **Sign in or paste a key**: Approve Spiral in the browser, or give the agent a key from the console.

### Use the Spiral CLI

1. **Install and sign in**: Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

```bash
npx -y spiral-cli@latest init
```

### Call the Spiral API

1. **Create an API key**: In the console, under API keys.

2. **Run Gemini 4 Argon**: One request; the result comes back by polling or a webhook.

```bash
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-argon",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Amp finds Gemini 4 Argon on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/amp

## Gemini 4 Argon tools and prices

- [Gemini 4 Argon](https://spiral.ai/tools/vertex/google/gemini-4-argon) (Vertex AI): 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

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| seed |  | — |  |
| stop |  | — |  |
| tools |  | — |  |
| reasoning |  | — | Reasoning effort; reasoning tokens count toward max_tokens. |
| max_tokens | 1 to 65536 | — | The most output tokens to generate. |
| tool_choice |  | — |  |

## What can you do with Gemini 4 Argon in Amp?

Gemini 4 Argon can help trace bugs, plan refactors, and investigate performance across a project's files. Give it repository conventions and measurable acceptance criteria so you can assess the proposed changes.

### Debug problems across multiple files

Ask Gemini 4 Argon to trace a failure through callers, state changes, and tests, then propose a fix with a regression test.

### Investigate performance bottlenecks

Use Gemini 4 Argon to examine slow code paths, establish a baseline, and compare proposed changes with measured results.

### Plan architectural changes

Give Gemini 4 Argon the current design and constraints. Request a migration plan that identifies dependencies, compatibility risks, and validation steps.

### Compare contract language

Ask Gemini 4 Argon to compare supplied clauses, quote the relevant passages, and list changes that need legal review.

### Explain financial results

Have Gemini 4 Argon compare supplied reporting periods, distinguish reported figures from calculations, and identify questions the evidence leaves unanswered.

### Review scientific arguments

Use Gemini 4 Argon to examine a supplied derivation or research argument, identify assumptions, and suggest checks for unsupported steps.

### Draft evidence-based technical briefs

Ask Gemini 4 Argon to synthesize supplied technical material into a brief with source references, disagreements, and open questions.

## Overview

- [Google Vertex AI](https://spiral.ai/providers/vertex)
- [All Amp guides](https://spiral.ai/integrations/amp)

## Gemini 4 Argon in other agents

- [ChatGPT](https://spiral.ai/integrations/chatgpt/gemini-4-argon)
- [Claude](https://spiral.ai/integrations/claude-desktop/gemini-4-argon)
- [Claude Code](https://spiral.ai/integrations/claude-code/gemini-4-argon)
- [Cline](https://spiral.ai/integrations/cline/gemini-4-argon)
- [Codex](https://spiral.ai/integrations/codex/gemini-4-argon)
- [Cursor](https://spiral.ai/integrations/cursor/gemini-4-argon)
- [Devin](https://spiral.ai/integrations/devin/gemini-4-argon)
- [dots](https://spiral.ai/integrations/dots/gemini-4-argon)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/gemini-4-argon)
- [Gemini](https://spiral.ai/integrations/gemini/gemini-4-argon)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/gemini-4-argon)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/gemini-4-argon)
- [goose](https://spiral.ai/integrations/goose/gemini-4-argon)
- [Grok](https://spiral.ai/integrations/grok/gemini-4-argon)
- [Grok Build](https://spiral.ai/integrations/grok-build/gemini-4-argon)
- [Hermes](https://spiral.ai/integrations/hermes/gemini-4-argon)
- [Instinct](https://spiral.ai/integrations/instinct/gemini-4-argon)
- [Manus](https://spiral.ai/integrations/manus/gemini-4-argon)
- [Muse](https://spiral.ai/integrations/muse/gemini-4-argon)
- [Muse Code](https://spiral.ai/integrations/muse-code/gemini-4-argon)
- [n8n](https://spiral.ai/integrations/n8n/gemini-4-argon)
- [Notion](https://spiral.ai/integrations/notion/gemini-4-argon)
- [OpenClaw](https://spiral.ai/integrations/openclaw/gemini-4-argon)
- [opencode](https://spiral.ai/integrations/opencode/gemini-4-argon)
- [Perplexity](https://spiral.ai/integrations/perplexity/gemini-4-argon)
- [Raycast](https://spiral.ai/integrations/raycast/gemini-4-argon)
- [VS Code](https://spiral.ai/integrations/vscode/gemini-4-argon)
- [Windsurf](https://spiral.ai/integrations/windsurf/gemini-4-argon)
- [Zapier](https://spiral.ai/integrations/zapier/gemini-4-argon)

## More Amp guides

- [Gemini 3.8 Flash](https://spiral.ai/integrations/amp/gemini-3-8-flash)
- [Gemini 4 Carbon](https://spiral.ai/integrations/amp/gemini-4-carbon)
- [Gemini 3.7 Flash](https://spiral.ai/integrations/amp/gemini-3-7-flash)
- [Gemini 4 Carbon Batch](https://spiral.ai/integrations/amp/gemini-4-carbon-batch)
- [Gemini 3.5 Flash-Lite](https://spiral.ai/integrations/amp/gemini-3-5-flash-lite)
- [Gemini 4 Hydrogen](https://spiral.ai/integrations/amp/gemini-4-hydrogen)

## Questions about Gemini 4 Argon in Amp

**Can Amp use Gemini 4 Argon?**

Yes. Add Spiral to Amp as an MCP server, and Amp can find and run Gemini 4 Argon like any built-in tool.

**How much does Gemini 4 Argon cost in Amp?**

Up to $0.106678 per run at the default inputs. $5.20 per 1M input tokens, $0.26 per 1M cached input tokens, $0.4875 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 Amp plan is separate.

**Do I need a Vertex AI account to use Gemini 4 Argon in Amp?**

No. Spiral runs Gemini 4 Argon 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 Argon in Amp?**

About two minutes: add the server, then sign in, then check it.

**Does Amp wait while Gemini 4 Argon runs?**

Gemini 4 Argon runs in the background. Amp starts the run, checks on it and gives you the text, json when it's ready.

**What is Gemini 4 Argon good at?**

Gemini 4 Argon is designed for complex software engineering, technical investigations, and analysis in finance, legal work, cybersecurity, and science. Useful tasks include reviewing code, comparing supplied documents, and checking mathematical arguments.

**How much context can Gemini 4 Argon handle?**

Its 1,048,576-token window includes input, thinking, function calls, and the final answer. Leave room for reasoning and the response when supplying a large amount of text.

**When should I use high reasoning?**

Google recommends high thinking for work across multiple code files, formal proofs, architecture design, and vulnerability discovery. Evaluate the available levels against representative tasks and keep the lowest level that produces acceptable results.

**Can Gemini 4 Argon return structured JSON?**

Yes, Gemini 4 Argon supports structured output. Define the required structure separately from the reasoning depth, and state how missing information should be represented.

**Does Gemini 4 Argon know current events?**

Its stated knowledge cutoff is March 2026, and some subjects may be less current. Supply dated source material when the answer depends on recent events or changes.

