[Integrations](https://spiral.ai/integrations) › [Instinct](https://spiral.ai/integrations/instinct) › [Magnum v4 72B](https://spiral.ai/integrations/instinct/magnum-v4-72b)

# How to Use Magnum v4 72B in Instinct

Magnum v4 72B targets creative prose and character roleplay. It fine-tunes Qwen2.5-72B-Instruct toward Claude 3 Sonnet and Opus prose, giving it a specific writing focus.

|  |  |
| --- | --- |
| Model | Magnum v4 72B |
| Available through | OpenRouter |
| Category | LLM |
| Price | Up to $0.026739 per run at the default inputs. $3.25 per 1M input tokens, $6.50 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, with a Spiral API key |

## Prompts to try with Magnum v4 72B in Instinct

- Each morning, use Magnum v4 72B to draft a 300-word mystery scene set in a railway station. Introduce one suspicious object and end before its meaning is explained.
- Every Monday, use Magnum v4 72B to prepare one tabletop NPC: a motive, a secret, three lines of dialogue, and two reactions to player questions. Keep the result under 500 words.
- Each Friday, use Magnum v4 72B to draft a 400-word scene about a traveling repair shop. Make the conflict practical and end with an unresolved choice.
- Every evening, use Magnum v4 72B to create a 250-word dialogue exercise between strangers who both claim the same suitcase. Keep their voices distinct and leave the dispute unresolved.
- Every Sunday, use Magnum v4 72B to write three 150-word story openings about an abandoned observatory. Use a temperature of 0.7 and give each opening a different viewpoint.

## How to set up Magnum v4 72B in Instinct

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

### Use the Spiral skill

1. **Send Instinct 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 Spiral's MCP server

1. **Create an API key**: In the console, open API keys and create a key for Instinct. Copy it.

2. **Send Instinct this message**: Instinct reads Spiral's skill and asks for your API key.

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

3. **Give Instinct the key**: When Instinct asks for the key, add it where Instinct keeps secrets, never in the chat itself.

### 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 Magnum v4 72B**: 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": "anthracite-org/magnum-v4-72b",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Instinct finds Magnum v4 72B on Spiral, shows the price and runs it.

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

## Magnum v4 72B tools and prices

- [Magnum v4 72B](https://spiral.ai/tools/openrouter/anthracite-org/magnum-v4-72b) (OpenRouter): Up to $0.026739 per run at the default inputs. $3.25 per 1M input tokens, $6.50 per 1M output tokens.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| seed |  | — |  |
| stop |  | — |  |
| min_p | 0 to 1 | — |  |
| top_k | 0 to … | — |  |
| top_p | 0 to 1 | — |  |
| logprobs |  | — |  |

## What can you do with Magnum v4 72B in Instinct?

Prepare recurring fiction drafts with Magnum v4 72B. Give each run a bounded scene brief and a clear ending.

### Game character dialogue

Draft a suspicious innkeeper's greetings, evasions, and confession.

### Interactive fiction

Write a scene that pauses for the player's next decision.

### Short story drafts

Turn a premise into a scene with a clear viewpoint.

### Dialogue revisions

Rewrite an exchange so each speaker has a distinct voice.

### Tabletop encounters

Prepare an NPC's motives, dialogue, and reactions to player choices.

### Serialized fiction

Draft the next scene from a continuity brief and unresolved conflict.

## Overview

- [OpenRouter](https://spiral.ai/providers/openrouter)
- [All Instinct guides](https://spiral.ai/integrations/instinct)

## Magnum v4 72B in other agents

- [Amp](https://spiral.ai/integrations/amp/magnum-v4-72b)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/magnum-v4-72b)
- [Claude](https://spiral.ai/integrations/claude-desktop/magnum-v4-72b)
- [Claude Code](https://spiral.ai/integrations/claude-code/magnum-v4-72b)
- [Cline](https://spiral.ai/integrations/cline/magnum-v4-72b)
- [Codex](https://spiral.ai/integrations/codex/magnum-v4-72b)
- [Cursor](https://spiral.ai/integrations/cursor/magnum-v4-72b)
- [Devin](https://spiral.ai/integrations/devin/magnum-v4-72b)
- [dots](https://spiral.ai/integrations/dots/magnum-v4-72b)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/magnum-v4-72b)
- [Gemini](https://spiral.ai/integrations/gemini/magnum-v4-72b)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/magnum-v4-72b)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/magnum-v4-72b)
- [goose](https://spiral.ai/integrations/goose/magnum-v4-72b)
- [Grok](https://spiral.ai/integrations/grok/magnum-v4-72b)
- [Grok Build](https://spiral.ai/integrations/grok-build/magnum-v4-72b)
- [Hermes](https://spiral.ai/integrations/hermes/magnum-v4-72b)
- [Manus](https://spiral.ai/integrations/manus/magnum-v4-72b)
- [Muse](https://spiral.ai/integrations/muse/magnum-v4-72b)
- [Muse Code](https://spiral.ai/integrations/muse-code/magnum-v4-72b)
- [n8n](https://spiral.ai/integrations/n8n/magnum-v4-72b)
- [Notion](https://spiral.ai/integrations/notion/magnum-v4-72b)
- [OpenClaw](https://spiral.ai/integrations/openclaw/magnum-v4-72b)
- [opencode](https://spiral.ai/integrations/opencode/magnum-v4-72b)
- [Perplexity](https://spiral.ai/integrations/perplexity/magnum-v4-72b)
- [Raycast](https://spiral.ai/integrations/raycast/magnum-v4-72b)
- [VS Code](https://spiral.ai/integrations/vscode/magnum-v4-72b)
- [Windsurf](https://spiral.ai/integrations/windsurf/magnum-v4-72b)
- [Zapier](https://spiral.ai/integrations/zapier/magnum-v4-72b)

## More Instinct guides

- [LongCat 2.0](https://spiral.ai/integrations/instinct/longcat-2-0)
- [Mi:dm K 3.0 Pro](https://spiral.ai/integrations/instinct/mi-dm-k-3-0-pro)
- [Llama Guard 4 12B](https://spiral.ai/integrations/instinct/llama-guard-4-12b)
- [MiMo V2.6 Pro](https://spiral.ai/integrations/instinct/mimo-v2-6-pro)
- [Llama 4 Scout](https://spiral.ai/integrations/instinct/llama-4-scout)
- [MiMo V2.6 Pro UltraSpeed](https://spiral.ai/integrations/instinct/mimo-v2-6-pro-ultraspeed)

## Questions about Magnum v4 72B in Instinct

**Can Instinct use Magnum v4 72B?**

Yes. Add Spiral to Instinct as an MCP server, and Instinct can find and run Magnum v4 72B like any built-in tool.

**How much does Magnum v4 72B cost in Instinct?**

Up to $0.026739 per run at the default inputs. $3.25 per 1M input tokens, $6.50 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Instinct plan is separate.

**Do I need an OpenRouter account to use Magnum v4 72B in Instinct?**

You need no OpenRouter account. Instinct uses one Spiral API key for every tool, and Spiral runs Magnum v4 72B on its own OpenRouter account.

**How long does it take to set up Magnum v4 72B in Instinct?**

About two minutes: create an API key, then send Instinct this message, then give Instinct the key.

**Does Instinct wait while Magnum v4 72B runs?**

Magnum v4 72B runs in the background. Instinct starts the run, checks on it and gives you the text, json when it's ready.

**What should I use Magnum v4 72B for?**

Start with fiction, dialogue, and character roleplay. Its stated focus is prose quality.

**Is Magnum v4 72B based on Claude?**

Its base is Qwen2.5-72B-Instruct. Claude 3 Sonnet and Opus prose provided the writing target.

**How should I prompt Magnum v4 72B for roleplay?**

Specify the character, situation, and voice. Ask it to leave your character's decisions and dialogue to you.

**How much context does Magnum v4 72B document?**

Its published configuration specifies 32,768 positions. For extended stories, keep a short continuity brief alongside the current scene.

**Which sampling settings should I try first?**

Try the maker's published values: temperature 0.7, top-k 20, and top-p 0.8. Compare drafts using the same scene before changing one setting at a time.

