[Integrations](https://spiral.ai/integrations) › [Claude Code](https://spiral.ai/integrations/claude-code) › [Gemini 3.5 Flash-Lite](https://spiral.ai/integrations/claude-code/gemini-3-5-flash-lite)

# How to Use Gemini 3.5 Flash-Lite in Claude Code

Gemini 3.5 Flash-Lite suits document extraction, classification, translation, and small coding tasks. Use minimal thinking for routine sorting and medium or high thinking for work that takes several tool calls.

|  |  |
| --- | --- |
| Model | Gemini 3.5 Flash-Lite |
| Available through | Vertex AI |
| Category | LLM |
| Price | Up to $0.013327 per run at the default inputs. $0.39 per 1M input tokens, $0.039 per 1M cached input tokens, $0.108333335 per 1M cache write tokens, $3.25 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Gemini 3.5 Flash-Lite in Claude Code

- Use Gemini 3.5 Flash-Lite with medium thinking to inspect the project's date parser, fix its handling of empty strings, and run the relevant tests. Keep the change limited to that behavior and report the test result.
- Use Gemini 3.5 Flash-Lite with minimal thinking to classify the issue descriptions in this project as bug, feature request, documentation, or unclear. Return JSON with the issue ID, label, and a supporting excerpt.
- Use Gemini 3.5 Flash-Lite with minimal thinking to extract configuration names and their documented defaults from the project's README. Return JSON and use null whenever a default is not stated.
- Use Gemini 3.5 Flash-Lite with medium thinking to add a small JavaScript utility that groups records by status. Follow the project's conventions and verify empty input and missing status values.
- Use Gemini 3.5 Flash-Lite to summarize the current diff in five bullets. Describe only changes visible in the diff and flag any behavior that needs a focused test.

## How to set up Gemini 3.5 Flash-Lite in Claude Code

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

### Use Spiral's MCP server

1. **Add the hosted server**: This adds Spiral to the current project. Add `--scope user` to use it in every project.

```bash
claude mcp add --transport http spiral https://api.spiral.ai/mcp
```

2. **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

1. **Send Claude Code 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 3.5 Flash-Lite**: 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-3.5-flash-lite",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Claude Code finds Gemini 3.5 Flash-Lite on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/claude-code

## Gemini 3.5 Flash-Lite tools and prices

- [Gemini 3.5 Flash Lite](https://spiral.ai/tools/vertex/google/gemini-3.5-flash-lite) (Vertex AI): Up to $0.013327 per run at the default inputs. $0.39 per 1M input tokens, $0.039 per 1M cached input tokens, $0.108333335 per 1M cache write tokens, $3.25 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 3.5 Flash-Lite in Claude Code?

Turn project text into structured records and handle narrowly scoped code fixes with Gemini 3.5 Flash-Lite. Use medium or high thinking when the work includes inspecting code, running commands, and checking results.

### Route support requests

Classify incoming messages as billing, access, bug, or other. Require an evidence excerpt and send ambiguous cases for review.

### Extract document records

Turn supplied purchase-order text into records containing order numbers, dates, quantities, and delivery addresses. Preserve source wording and leave missing values empty.

### Translate customer messages

Translate support replies while preserving product names, order numbers, and supplied terminology. Ask for a separate list of ambiguous phrases.

### Summarize operational updates

Condense supplied status updates into completed work, blockers, and next actions. Include a supporting excerpt for each item.

### Make small code fixes

Handle a narrow bug fix or utility function with clear acceptance criteria. Use medium or high thinking when the work includes running tests and inspecting results.

### Tag searchable content

Assign approved topic labels to articles, help-center entries, or product descriptions. Provide examples and allow an unclassified result.

### Answer from supplied documentation

Answer a focused question using supplied manuals or policy text. Require supporting passages and an explicit answer when the material does not contain the information.

## Overview

- [Google Vertex AI](https://spiral.ai/providers/vertex)
- [All Claude Code guides](https://spiral.ai/integrations/claude-code)

## Gemini 3.5 Flash-Lite in other agents

- [Amp](https://spiral.ai/integrations/amp/gemini-3-5-flash-lite)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/gemini-3-5-flash-lite)
- [Claude](https://spiral.ai/integrations/claude-desktop/gemini-3-5-flash-lite)
- [Cline](https://spiral.ai/integrations/cline/gemini-3-5-flash-lite)
- [Codex](https://spiral.ai/integrations/codex/gemini-3-5-flash-lite)
- [Cursor](https://spiral.ai/integrations/cursor/gemini-3-5-flash-lite)
- [Devin](https://spiral.ai/integrations/devin/gemini-3-5-flash-lite)
- [dots](https://spiral.ai/integrations/dots/gemini-3-5-flash-lite)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/gemini-3-5-flash-lite)
- [Gemini](https://spiral.ai/integrations/gemini/gemini-3-5-flash-lite)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/gemini-3-5-flash-lite)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/gemini-3-5-flash-lite)
- [goose](https://spiral.ai/integrations/goose/gemini-3-5-flash-lite)
- [Grok](https://spiral.ai/integrations/grok/gemini-3-5-flash-lite)
- [Grok Build](https://spiral.ai/integrations/grok-build/gemini-3-5-flash-lite)
- [Hermes](https://spiral.ai/integrations/hermes/gemini-3-5-flash-lite)
- [Instinct](https://spiral.ai/integrations/instinct/gemini-3-5-flash-lite)
- [Manus](https://spiral.ai/integrations/manus/gemini-3-5-flash-lite)
- [Muse](https://spiral.ai/integrations/muse/gemini-3-5-flash-lite)
- [Muse Code](https://spiral.ai/integrations/muse-code/gemini-3-5-flash-lite)
- [n8n](https://spiral.ai/integrations/n8n/gemini-3-5-flash-lite)
- [Notion](https://spiral.ai/integrations/notion/gemini-3-5-flash-lite)
- [OpenClaw](https://spiral.ai/integrations/openclaw/gemini-3-5-flash-lite)
- [opencode](https://spiral.ai/integrations/opencode/gemini-3-5-flash-lite)
- [Perplexity](https://spiral.ai/integrations/perplexity/gemini-3-5-flash-lite)
- [Raycast](https://spiral.ai/integrations/raycast/gemini-3-5-flash-lite)
- [VS Code](https://spiral.ai/integrations/vscode/gemini-3-5-flash-lite)
- [Windsurf](https://spiral.ai/integrations/windsurf/gemini-3-5-flash-lite)
- [Zapier](https://spiral.ai/integrations/zapier/gemini-3-5-flash-lite)

## More Claude Code guides

- [Gemini 3.5 Flash](https://spiral.ai/integrations/claude-code/gemini-3-5-flash)
- [Gemini 3.7 Flash](https://spiral.ai/integrations/claude-code/gemini-3-7-flash)
- [Gemini 3.1 Pro Preview](https://spiral.ai/integrations/claude-code/gemini-3-1-pro-preview)
- [Gemini 3.8 Flash](https://spiral.ai/integrations/claude-code/gemini-3-8-flash)
- [Gemini 3.1 Pro Custom Tools](https://spiral.ai/integrations/claude-code/gemini-3-1-pro-custom-tools)
- [Gemini 4 Argon](https://spiral.ai/integrations/claude-code/gemini-4-argon)

## Questions about Gemini 3.5 Flash-Lite in Claude Code

**Can Claude Code use Gemini 3.5 Flash-Lite?**

Yes. Add Spiral to Claude Code as an MCP server, and Claude Code can find and run Gemini 3.5 Flash-Lite like any built-in tool.

**How much does Gemini 3.5 Flash-Lite cost in Claude Code?**

Up to $0.013327 per run at the default inputs. $0.39 per 1M input tokens, $0.039 per 1M cached input tokens, $0.108333335 per 1M cache write tokens, $3.25 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 a Vertex AI account to use Gemini 3.5 Flash-Lite in Claude Code?**

No. Spiral runs Gemini 3.5 Flash-Lite 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 3.5 Flash-Lite in Claude Code?**

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

**Does Claude Code wait while Gemini 3.5 Flash-Lite runs?**

Gemini 3.5 Flash-Lite runs in the background. Claude Code starts the run, checks on it and gives you the text, json when it's ready.

**What is Gemini 3.5 Flash-Lite good at?**

Use it for document extraction, classification, translation, and small coding tasks. These are useful starting points when you have many clearly defined jobs.

**How much text can Gemini 3.5 Flash-Lite handle?**

Google lists an input limit of 1,048,576 tokens and an output limit of 65,536 tokens. Treat those as ceilings and request only the output your task needs.

**When should I increase its thinking level?**

Keep minimal thinking for routine extraction, classification, and routing. Use medium or high thinking for autonomous coding or several tool calls, since minimal thinking can end multi-step work prematurely.

**Can Gemini 3.5 Flash-Lite return JSON?**

Yes. Give it a defined schema for extraction or classification, then check the returned values against the source because the model can still hallucinate.

**How should I prompt Gemini 3.5 Flash-Lite?**

Define the task, required format, and what to do with missing information. Include a few examples; for lengthy source material, place your specific question at the end.

