[Integrations](https://spiral.ai/integrations) › [Gemini CLI](https://spiral.ai/integrations/gemini-cli) › [Ministral 3 14B](https://spiral.ai/integrations/gemini-cli/ministral-3-14b)

# How to Use Ministral 3 14B in Gemini CLI

Ministral 3 14B handles instruction following, multilingual chat, and structured extraction. Unlike Ministral 3 14B Reasoning, it is the instruction-tuned choice for direct responses rather than extended reasoning.

|  |  |
| --- | --- |
| Model | Ministral 3 14B |
| Available through | OpenRouter |
| Category | LLM |
| Price | Up to $0.001076 per run at the default inputs. $0.26 per 1M input tokens, $0.26 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Ministral 3 14B in Gemini CLI

- Use Ministral 3 14B to read the selected module and draft documentation for its exported functions. Include parameters, return values, and one usage example; flag behavior the code does not establish.
- Use Ministral 3 14B to classify the issues in this project's issue export as bug, feature request, documentation, or unclear. Return JSON with each issue ID, category, and a supporting excerpt.
- Use Ministral 3 14B to compare the README with the selected implementation files. List each conflicting statement, the relevant file location, and a suggested correction.
- Use Ministral 3 14B to translate the project's selected help text into French and Spanish. Preserve interpolation tokens, code examples, and product names; follow the existing glossary.
- Use Ministral 3 14B with temperature 0.05 to extract configuration requirements from this project's documentation. Return JSON containing each requirement, its source file, and an exact supporting excerpt.

## How to set up Ministral 3 14B 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 Ministral 3 14B**: 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": "mistralai/ministral-14b-2512",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Gemini CLI finds Ministral 3 14B on Spiral, shows the price and runs it.

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

## Ministral 3 14B tools and prices

- [Mistral: Ministral 3 14B 2512](https://spiral.ai/tools/openrouter/mistralai/ministral-14b-2512) (OpenRouter): Up to $0.001076 per run at the default inputs. $0.26 per 1M input tokens, $0.26 per 1M output tokens.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| seed |  | — |  |
| stop |  | — |  |
| tools |  | — |  |
| top_p | 0 to 1 | — |  |
| max_tokens | 1 to 209715 | — | The most output tokens to generate. |
| temperature | 0 to 2 | — |  |

## What can you do with Ministral 3 14B in Gemini CLI?

Turn selected project files into structured explanations, documentation drafts, and review checklists with Ministral 3 14B. Give the agent a focused task and an example of the output you want.

### Support ticket classification

Assign incoming messages to defined categories and extract the customer's request into JSON. Include an example for ambiguous tickets.

### Multilingual reply drafts

Draft replies in the customer's requested language using supplied policies. Preserve product names and flag questions the policy does not answer.

### Meeting action lists

Turn meeting transcripts into decisions, owners, and deadlines. Require supporting excerpts and leave unstated deadlines empty.

### Document question answering

Ask focused questions about supplied manuals or reports. Request a supporting passage for each answer and an explicit response when evidence is missing.

### Repository documentation

Have a coding agent turn selected source files into function descriptions, usage examples, and a list of undocumented behavior to check.

### Scheduled change digests

Compare newly retrieved announcements with the previous run and produce a concise list of changes with source links.

### Structured record extraction

Extract names, dates, identifiers, and stated requirements from supplied text. Define the output structure and show how missing values should appear.

### Tool-assisted status answers

Let an agent retrieve a record, then explain its status using only the returned data. Specify when it should request more information.

## Overview

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

## Ministral 3 14B in other agents

- [Amp](https://spiral.ai/integrations/amp/ministral-3-14b)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/ministral-3-14b)
- [Claude](https://spiral.ai/integrations/claude-desktop/ministral-3-14b)
- [Claude Code](https://spiral.ai/integrations/claude-code/ministral-3-14b)
- [Cline](https://spiral.ai/integrations/cline/ministral-3-14b)
- [Codex](https://spiral.ai/integrations/codex/ministral-3-14b)
- [Cursor](https://spiral.ai/integrations/cursor/ministral-3-14b)
- [Devin](https://spiral.ai/integrations/devin/ministral-3-14b)
- [dots](https://spiral.ai/integrations/dots/ministral-3-14b)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/ministral-3-14b)
- [Gemini](https://spiral.ai/integrations/gemini/ministral-3-14b)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/ministral-3-14b)
- [goose](https://spiral.ai/integrations/goose/ministral-3-14b)
- [Grok](https://spiral.ai/integrations/grok/ministral-3-14b)
- [Grok Build](https://spiral.ai/integrations/grok-build/ministral-3-14b)
- [Hermes](https://spiral.ai/integrations/hermes/ministral-3-14b)
- [Instinct](https://spiral.ai/integrations/instinct/ministral-3-14b)
- [Manus](https://spiral.ai/integrations/manus/ministral-3-14b)
- [Muse](https://spiral.ai/integrations/muse/ministral-3-14b)
- [Muse Code](https://spiral.ai/integrations/muse-code/ministral-3-14b)
- [n8n](https://spiral.ai/integrations/n8n/ministral-3-14b)
- [Notion](https://spiral.ai/integrations/notion/ministral-3-14b)
- [OpenClaw](https://spiral.ai/integrations/openclaw/ministral-3-14b)
- [opencode](https://spiral.ai/integrations/opencode/ministral-3-14b)
- [Perplexity](https://spiral.ai/integrations/perplexity/ministral-3-14b)
- [Raycast](https://spiral.ai/integrations/raycast/ministral-3-14b)
- [VS Code](https://spiral.ai/integrations/vscode/ministral-3-14b)
- [Windsurf](https://spiral.ai/integrations/windsurf/ministral-3-14b)
- [Zapier](https://spiral.ai/integrations/zapier/ministral-3-14b)

## More Gemini CLI guides

- [MiniMax M3.1 Flash](https://spiral.ai/integrations/gemini-cli/minimax-m3-1-flash)
- [Ministral 3 3B](https://spiral.ai/integrations/gemini-cli/ministral-3-3b)
- [MiniMax M3 (batch)](https://spiral.ai/integrations/gemini-cli/minimax-m3-batch)
- [Ministral 3 8B](https://spiral.ai/integrations/gemini-cli/ministral-3-8b)
- [MiniMax M3](https://spiral.ai/integrations/gemini-cli/minimax-m3)
- [Mistral Large](https://spiral.ai/integrations/gemini-cli/mistral-large)

## Questions about Ministral 3 14B in Gemini CLI

**Can Gemini CLI use Ministral 3 14B?**

Yes. Add Spiral to Gemini CLI as an MCP server, and Gemini CLI can find and run Ministral 3 14B like any built-in tool.

**How much does Ministral 3 14B cost in Gemini CLI?**

Up to $0.001076 per run at the default inputs. $0.26 per 1M input tokens, $0.26 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 Ministral 3 14B in Gemini CLI?**

No. Spiral runs Ministral 3 14B on its own OpenRouter account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Ministral 3 14B in Gemini CLI?**

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

**Does Gemini CLI wait while Ministral 3 14B runs?**

Ministral 3 14B runs in the background. Gemini CLI starts the run, checks on it and gives you the text, json when it's ready.

**What is Ministral 3 14B good for?**

Use it for instruction following, multilingual conversation, and extracting information into JSON. Give it a defined task and an example of the result you want.

**How much context can Ministral 3 14B handle?**

Mistral specifies a 256k context window, or 262,144 tokens. Treat that as context capacity, not a promise of an equally long answer.

**Does Ministral 3 14B support languages besides English?**

Yes. Mistral describes Ministral 3 as supporting more than 40 native languages; specify the response language and provide examples of any terminology you need preserved.

**Is Ministral 3 14B the same as Ministral 3 14B Reasoning?**

No. Ministral 3 14B is instruction-tuned, while Ministral 3 14B Reasoning is separately trained to think longer on reasoning tasks. Keep that distinction in mind when reading benchmark results.

**What temperature should I use with Ministral 3 14B?**

Mistral recommends a temperature below 0.1 for routine and production work. Start there for extraction and classification, then compare results on representative examples.

