[Integrations](https://spiral.ai/integrations) › [Muse](https://spiral.ai/integrations/muse) › [Qwen2.5 Coder 32B Instruct](https://spiral.ai/integrations/muse/qwen2-5-coder-32b-instruct)

# How to Use Qwen2.5 Coder 32B Instruct in Muse

Qwen2.5 Coder 32B Instruct writes, explains, and repairs code from instructions. It scored above Qwen2.5 Coder 14B Instruct on Qwen's Aider editing test, making testable code changes a practical use.

|  |  |
| --- | --- |
| Model | Qwen2.5 Coder 32B Instruct |
| Available through | OpenRouter |
| Category | LLM |
| Price | Up to $0.005356 per run at the default inputs. $0.858 per 1M input tokens, $1.30 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, with a Spiral API key |

## Prompts to try with Qwen2.5 Coder 32B Instruct in Muse

- Use Qwen2.5 Coder 32B Instruct to write a Python 3.11 function that merges overlapping integer intervals. Treat touching endpoints as overlapping, return sorted intervals, and include tests for empty input and nested intervals.
- Use Qwen2.5 Coder 32B Instruct to explain why this Python function returns 3: def count(xs): return sum(x > 0 for x in xs). Trace the input [-2, 1, 0, 4, 7] and show each intermediate Boolean value.
- Use Qwen2.5 Coder 32B Instruct to fix this JavaScript function: function firstPositive(xs) { return xs.find(x => x > 0) || null; }. The intended result is the first nonnegative number, including zero. Include tests that expose the current mistakes.
- Use Qwen2.5 Coder 32B Instruct to translate this Python routine into TypeScript: def unique_sorted(xs): return sorted(set(xs)). Accept numbers only, preserve ascending numeric order, and include examples with duplicates and negative values.
- Use Qwen2.5 Coder 32B Instruct to write PostgreSQL for orders(id, customer_id, status, total). Return each customer's completed-order revenue, including customers with no completed orders from customers(id, name). Explain the join and include a small worked example.

## How to set up Qwen2.5 Coder 32B Instruct in Muse

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

### Use the Muse connector

1. **Create an API key**: Create one under API keys in the Spiral console.

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

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

3. **Give Muse the key**: Paste the key when Muse asks for it, never in the chat itself.

### Use the Spiral skill

1. **Send Muse 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 Qwen2.5 Coder 32B 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": "qwen/qwen-2.5-coder-32b-instruct",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Muse finds Qwen2.5 Coder 32B Instruct on Spiral, shows the price and runs it.

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

## Qwen2.5 Coder 32B Instruct tools and prices

- [Qwen2.5 Coder 32B Instruct](https://spiral.ai/tools/openrouter/qwen/qwen-2.5-coder-32b-instruct) (OpenRouter): Up to $0.005356 per run at the default inputs. $0.858 per 1M input tokens, $1.30 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 | — |  |
| max_tokens | 1 to 29491 | — | The most output tokens to generate. |

## What can you do with Qwen2.5 Coder 32B Instruct in Muse?

Turn a coding question into an implementation, execution trace, or proposed repair with Qwen2.5 Coder. Paste the relevant code and expected behavior so the conversation stays focused on a checkable result.

### Repair a reproducible bug

Give Qwen2.5 Coder the failing code, error message, and expected behavior. Request a small patch and a regression test that demonstrates the fix.

### Turn requirements into a function

Specify the language, function signature, allowed dependencies, and sample inputs and outputs. Ask for an implementation with tests for empty inputs and boundary cases.

### Explain unfamiliar code

Ask Qwen2.5 Coder to trace a function using concrete inputs, identify state changes, and explain the returned value. Use the explanation to guide a review or debugging session.

### Draft regression tests

Provide a function and its intended behavior. Request tests for normal cases, invalid inputs, and the specific failure that prompted the change.

### Translate a small routine between languages

Provide the original routine and examples of its behavior. Request a translation that preserves those examples, with explicit notes about differences in types or error handling.

### Draft SQL for a reporting question

Supply the database dialect, table definitions, example rows, and reporting rules. Ask for a query and an explanation of joins, filters, and aggregation.

### Refactor a function with behavior checks

Give Qwen2.5 Coder a function and its tests. Request clearer control flow while preserving the interface, outputs, and error behavior.

## Overview

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

## Qwen2.5 Coder 32B Instruct in other agents

- [Amp](https://spiral.ai/integrations/amp/qwen2-5-coder-32b-instruct)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/qwen2-5-coder-32b-instruct)
- [Claude](https://spiral.ai/integrations/claude-desktop/qwen2-5-coder-32b-instruct)
- [Claude Code](https://spiral.ai/integrations/claude-code/qwen2-5-coder-32b-instruct)
- [Cline](https://spiral.ai/integrations/cline/qwen2-5-coder-32b-instruct)
- [Codex](https://spiral.ai/integrations/codex/qwen2-5-coder-32b-instruct)
- [Cursor](https://spiral.ai/integrations/cursor/qwen2-5-coder-32b-instruct)
- [Devin](https://spiral.ai/integrations/devin/qwen2-5-coder-32b-instruct)
- [dots](https://spiral.ai/integrations/dots/qwen2-5-coder-32b-instruct)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/qwen2-5-coder-32b-instruct)
- [Gemini](https://spiral.ai/integrations/gemini/qwen2-5-coder-32b-instruct)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/qwen2-5-coder-32b-instruct)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/qwen2-5-coder-32b-instruct)
- [goose](https://spiral.ai/integrations/goose/qwen2-5-coder-32b-instruct)
- [Grok](https://spiral.ai/integrations/grok/qwen2-5-coder-32b-instruct)
- [Grok Build](https://spiral.ai/integrations/grok-build/qwen2-5-coder-32b-instruct)
- [Hermes](https://spiral.ai/integrations/hermes/qwen2-5-coder-32b-instruct)
- [Instinct](https://spiral.ai/integrations/instinct/qwen2-5-coder-32b-instruct)
- [Manus](https://spiral.ai/integrations/manus/qwen2-5-coder-32b-instruct)
- [Muse Code](https://spiral.ai/integrations/muse-code/qwen2-5-coder-32b-instruct)
- [n8n](https://spiral.ai/integrations/n8n/qwen2-5-coder-32b-instruct)
- [Notion](https://spiral.ai/integrations/notion/qwen2-5-coder-32b-instruct)
- [OpenClaw](https://spiral.ai/integrations/openclaw/qwen2-5-coder-32b-instruct)
- [opencode](https://spiral.ai/integrations/opencode/qwen2-5-coder-32b-instruct)
- [Perplexity](https://spiral.ai/integrations/perplexity/qwen2-5-coder-32b-instruct)
- [Raycast](https://spiral.ai/integrations/raycast/qwen2-5-coder-32b-instruct)
- [VS Code](https://spiral.ai/integrations/vscode/qwen2-5-coder-32b-instruct)
- [Windsurf](https://spiral.ai/integrations/windsurf/qwen2-5-coder-32b-instruct)
- [Zapier](https://spiral.ai/integrations/zapier/qwen2-5-coder-32b-instruct)

## More Muse guides

- [Qwen2.5 72B Instruct](https://spiral.ai/integrations/muse/qwen2-5-72b-instruct)
- [Qwen2.5-VL 72B Instruct](https://spiral.ai/integrations/muse/qwen2-5-vl-72b-instruct)
- [Qwen Plus](https://spiral.ai/integrations/muse/qwen-plus)
- [Qwen3 14B](https://spiral.ai/integrations/muse/qwen3-14b)
- [Pixtral Large 2502](https://spiral.ai/integrations/muse/pixtral-large-2502)
- [Qwen3 235B A22B Instruct 2507](https://spiral.ai/integrations/muse/qwen3-235b-a22b-instruct-2507)

## Questions about Qwen2.5 Coder 32B Instruct in Muse

**Can Muse use Qwen2.5 Coder 32B Instruct?**

Yes. Add Spiral to Muse as an MCP server, and Muse can find and run Qwen2.5 Coder 32B Instruct like any built-in tool.

**How much does Qwen2.5 Coder 32B Instruct cost in Muse?**

Up to $0.005356 per run at the default inputs. $0.858 per 1M input tokens, $1.30 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Muse plan is separate.

**Do I need an OpenRouter account to use Qwen2.5 Coder 32B Instruct in Muse?**

You need no OpenRouter account. Muse uses one Spiral API key for every tool, and Spiral runs Qwen2.5 Coder 32B Instruct on its own OpenRouter account.

**How long does it take to set up Qwen2.5 Coder 32B Instruct in Muse?**

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

**Does Muse wait while Qwen2.5 Coder 32B Instruct runs?**

Qwen2.5 Coder 32B Instruct runs in the background. Muse starts the run, checks on it and gives you the text, json when it's ready.

**What is Qwen2.5 Coder good at?**

Use it for writing code, explaining execution, and proposing repairs. Give it a concrete task with expected behavior so you can check the result.

**Which programming languages does Qwen2.5 Coder support?**

Qwen reports evaluation across more than 40 programming languages, including Haskell and Racket. Test it on examples from your target language before relying on its suggestions.

**How much context can Qwen2.5 Coder handle?**

The model card specifies 32,768 tokens in the default configuration and extended support for 131,072 tokens with YaRN. Treat the larger figure as conditional on the model configuration, and keep requests focused on relevant code.

**Where does Qwen2.5 Coder need checking?**

Code repairs need verification: Qwen reports a 60.9% first-attempt score on its Aider editing evaluation. Ask for a regression test and run it before accepting a proposed fix.

**Which sampling settings should I start with?**

Qwen's published configuration uses temperature 0.7, nucleus sampling at 0.8, and a shortlist of 20 candidate tokens. Use those as a baseline, then compare results on representative coding tasks.

