[Integrations](https://spiral.ai/integrations) › [Gemini CLI](https://spiral.ai/integrations/gemini-cli) › [Gemma 2 27B Instruct](https://spiral.ai/integrations/gemini-cli/gemma-2-27b-instruct)

# How to Use Gemma 2 27B Instruct in Gemini CLI

Gemma 2 27B Instruct is for English summaries, questions about supplied text, and response drafts. It uses 27 billion parameters, compared with Gemma 2 9B Instruct's 9 billion, while retaining an 8,192-token context.

|  |  |
| --- | --- |
| Model | Gemma 2 27B Instruct |
| Available through | OpenRouter |
| Category | LLM |
| Price | Up to $0.001761 per run at the default inputs. $0.845 per 1M input tokens, $0.845 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Gemma 2 27B Instruct in Gemini CLI

- Use Gemma 2 27B Instruct to summarize the current diff and relevant README passages. Select excerpts that fit an 8,192-token context with room for the answer. Return a review brief under 300 words covering changed behavior, documented expectations, and unanswered questions.
- Use Gemma 2 27B Instruct to answer this question from the project's authentication documentation: when does a session expire? Send only relevant excerpts, cite their paths, and flag any missing information. Keep the answer under 200 words.
- Use Gemma 2 27B Instruct to compare the selected handler with its documented behavior. Provide compact code and documentation excerpts. Return a table of documented requirements, matching evidence, and questions for a reviewer; do not claim the code was tested.
- Use Gemma 2 27B Instruct to draft release notes from the current diff. Include only changes supported by the supplied excerpts, keep the draft under 250 words, and list uncertain user-facing effects separately.
- Use Gemma 2 27B Instruct to summarize the failing test output and the relevant test excerpt. Describe the observed failure, expected behavior, and information still needed in no more than 200 words.

## How to set up Gemma 2 27B Instruct in Gemini CLI

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

### Use Spiral's MCP server

1. **Add the server**: Or add it to `~/.gemini/settings.json` yourself:

```bash
gemini mcp add --scope user --transport http spiral https://api.spiral.ai/mcp
```

2. **Sign in**: Start `gemini` and run:

```text
/mcp auth spiral
```

3. **Check it**: In Gemini CLI, run `/mcp`. From your shell:

```bash
gemini mcp list
```

### Use the Spiral skill

1. **Send Gemini CLI 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 Gemma 2 27B Instruct**: 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": "openrouter",
    "model": "google/gemma-2-27b-it",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Gemini CLI finds Gemma 2 27B Instruct on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/gemini-cli

## Gemma 2 27B Instruct tools and prices

- [Google: Gemma 2 27B](https://spiral.ai/tools/openrouter/google/gemma-2-27b-it) (OpenRouter): Up to $0.001761 per run at the default inputs. $0.845 per 1M input tokens, $0.845 per 1M output tokens.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| seed |  | — |  |
| stop |  | — |  |
| top_p | 0 to 1 | — |  |
| max_tokens | 1 to 2048 | — | The most output tokens to generate. |
| temperature | 0 to 2 | — |  |
| response_format |  | — |  |

## What can you do with Gemma 2 27B Instruct in Gemini CLI?

Turn selected project text into review briefs, documentation summaries, and answers grounded in the code excerpts you provide. Give Gemma 2 a focused set of files or diffs that fits its context, and ask it to separate evidence from assumptions.

### Report summaries

Ask Gemma 2 to condense a report excerpt into findings, supporting evidence, and open questions. Request a short passage supporting each finding.

### Questions about policies

Give Gemma 2 the relevant policy passages and ask it to answer a specific employee or customer question. Require it to identify anything the passages leave unresolved.

### Customer reply drafts

Provide Gemma 2 with a customer conversation and approved guidance. Ask for a concise reply that addresses the complaint without inventing promises.

### Conversation sentiment review

Ask Gemma 2 to label the sentiment of each short conversation and quote the wording behind its assessment. Keep ambiguous cases visible for review.

### Project change summaries

Have Gemma 2 summarize selected diffs and related documentation into a review brief. Ask it to distinguish changes shown in the text from questions that need testing.

### Meeting recaps

Give Gemma 2 a short meeting transcript and request decisions, assigned actions, and unresolved questions. Tell it to leave owners and deadlines unspecified when the transcript does.

## Overview

- [OpenRouter](https://spiral.ai/providers/openrouter)
- [All Gemini CLI guides](https://spiral.ai/integrations/gemini-cli)

## Gemma 2 27B Instruct in other agents

- [Amp](https://spiral.ai/integrations/amp/gemma-2-27b-instruct)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/gemma-2-27b-instruct)
- [Claude](https://spiral.ai/integrations/claude-desktop/gemma-2-27b-instruct)
- [Claude Code](https://spiral.ai/integrations/claude-code/gemma-2-27b-instruct)
- [Cline](https://spiral.ai/integrations/cline/gemma-2-27b-instruct)
- [Codex](https://spiral.ai/integrations/codex/gemma-2-27b-instruct)
- [Cursor](https://spiral.ai/integrations/cursor/gemma-2-27b-instruct)
- [Devin](https://spiral.ai/integrations/devin/gemma-2-27b-instruct)
- [dots](https://spiral.ai/integrations/dots/gemma-2-27b-instruct)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/gemma-2-27b-instruct)
- [Gemini](https://spiral.ai/integrations/gemini/gemma-2-27b-instruct)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/gemma-2-27b-instruct)
- [goose](https://spiral.ai/integrations/goose/gemma-2-27b-instruct)
- [Grok](https://spiral.ai/integrations/grok/gemma-2-27b-instruct)
- [Grok Build](https://spiral.ai/integrations/grok-build/gemma-2-27b-instruct)
- [Hermes](https://spiral.ai/integrations/hermes/gemma-2-27b-instruct)
- [Instinct](https://spiral.ai/integrations/instinct/gemma-2-27b-instruct)
- [Manus](https://spiral.ai/integrations/manus/gemma-2-27b-instruct)
- [Muse](https://spiral.ai/integrations/muse/gemma-2-27b-instruct)
- [Muse Code](https://spiral.ai/integrations/muse-code/gemma-2-27b-instruct)
- [n8n](https://spiral.ai/integrations/n8n/gemma-2-27b-instruct)
- [Notion](https://spiral.ai/integrations/notion/gemma-2-27b-instruct)
- [OpenClaw](https://spiral.ai/integrations/openclaw/gemma-2-27b-instruct)
- [opencode](https://spiral.ai/integrations/opencode/gemma-2-27b-instruct)
- [Perplexity](https://spiral.ai/integrations/perplexity/gemma-2-27b-instruct)
- [Raycast](https://spiral.ai/integrations/raycast/gemma-2-27b-instruct)
- [VS Code](https://spiral.ai/integrations/vscode/gemma-2-27b-instruct)
- [Windsurf](https://spiral.ai/integrations/windsurf/gemma-2-27b-instruct)
- [Zapier](https://spiral.ai/integrations/zapier/gemma-2-27b-instruct)

## More Gemini CLI guides

- [Gemini Pro Latest](https://spiral.ai/integrations/gemini-cli/gemini-pro-latest)
- [Gemma 3 12B](https://spiral.ai/integrations/gemini-cli/gemma-3-12b)
- [Gemini Flash Latest](https://spiral.ai/integrations/gemini-cli/gemini-flash-latest)
- [Gemma 3 27B IT](https://spiral.ai/integrations/gemini-cli/gemma-3-27b-it)
- [Gemini 4 Neon](https://spiral.ai/integrations/gemini-cli/gemini-4-neon)
- [Gemma 3 4B](https://spiral.ai/integrations/gemini-cli/gemma-3-4b)

## Questions about Gemma 2 27B Instruct in Gemini CLI

**Can Gemini CLI use Gemma 2 27B Instruct?**

Yes. Add Spiral to Gemini CLI as an MCP server, and Gemini CLI can find and run Gemma 2 27B Instruct like any built-in tool.

**How much does Gemma 2 27B Instruct cost in Gemini CLI?**

Up to $0.001761 per run at the default inputs. $0.845 per 1M input tokens, $0.845 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Gemini CLI plan is separate.

**Do I need an OpenRouter account to use Gemma 2 27B Instruct in Gemini CLI?**

No. Spiral runs Gemma 2 27B Instruct on its own OpenRouter account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Gemma 2 27B Instruct in Gemini CLI?**

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

**Does Gemini CLI wait while Gemma 2 27B Instruct runs?**

Gemma 2 27B Instruct runs in the background. Gemini CLI starts the run, checks on it and gives you the text, json when it's ready.

**What is Gemma 2 useful for?**

Use Gemma 2 for English summaries, questions about supplied text, and reasoning over short excerpts. Google also demonstrates conversation sentiment analysis and reply drafting.

**How much text can Gemma 2 work with?**

Gemma 2 has an 8,192-token context. Keep instructions and source excerpts compact, leave room for the response, and split longer material into separate tasks.

**Can Gemma 2 understand images or audio?**

Gemma 2 is text-to-text. For questions about a recording or screenshot, supply a transcript or text description.

**Which languages should I use with Gemma 2?**

Start with English. Google says Gemma 2 was not trained specifically for leading multilingual performance, so test other languages on examples from your actual task.

**Can I trust Gemma 2's factual answers?**

Gemma 2 can generate incorrect or outdated statements. For factual work, provide the relevant source text, request supporting passages, and check the answer against them.

