[Integrations](https://spiral.ai/integrations) › [Devin](https://spiral.ai/integrations/devin) › [Qwen3.5-Flash](https://spiral.ai/integrations/devin/qwen3-5-flash)

# How to Use Qwen3.5-Flash in Devin

Qwen3.5-Flash suits document summaries, information extraction, and everyday coding tasks. It combines the model behind Qwen3.5-35B-A3B with a 1 million token context window for larger collections of source material.

|  |  |
| --- | --- |
| Model | Qwen3.5-Flash |
| Available through | OpenRouter |
| Category | LLM |
| Price | Up to $0.001389 per run at the default inputs. $0.0845 per 1M input tokens, $0.338 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Qwen3.5-Flash in Devin

- Use Qwen3.5-Flash to trace how this project's login request reaches session creation. Read the relevant source files and return a numbered execution path with file references and unresolved questions.
- Use Qwen3.5-Flash with thinking enabled to investigate the failing test output and related implementation. Propose the smallest patch, explain the failure mechanism, and identify a test that would distinguish the fix from the current behavior.
- Use Qwen3.5-Flash to compare the API documentation with the route handlers in this project. Return a table of conflicting request requirements, response shapes, and error behavior, with evidence from both.
- Use Qwen3.5-Flash with reasoning disabled to extract the project's documented environment variables as JSON. Include name, purpose, required status, and source path; use null when the documentation does not supply a value.
- Use Qwen3.5-Flash to review this diff and its surrounding code for changed behavior. Return only findings that include a concrete trigger, likely consequence, and relevant file location.

## How to set up Qwen3.5-Flash in Devin

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

### Use Spiral's MCP server

1. **Add a custom MCP**: Open Customize › MCPs. Click + and choose New MCP. Fill in:

2. **Sign in to Spiral and approve**: In Customize › MCPs, open Spiral and choose Connect. Spiral opens: sign in, pick the organization and click Authorize. No API key.

3. **Check it**: On Spiral's configuration page, click Test tools. Devin lists Spiral's tools.

### Use the Spiral skill

1. **Send Devin 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 Qwen3.5-Flash**: 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/qwen3.5-flash-02-23",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Devin finds Qwen3.5-Flash on Spiral, shows the price and runs it.

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

## Qwen3.5-Flash tools and prices

- [Qwen: Qwen3.5-Flash](https://spiral.ai/tools/openrouter/qwen/qwen3.5-flash-02-23) (OpenRouter): Up to $0.001389 per run at the default inputs. $0.0845 per 1M input tokens, $0.338 per 1M output tokens.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| seed |  | — |  |
| stop |  | — |  |
| tools |  | — |  |
| top_k | 0 to … | — |  |
| top_p | 0 to 1 | — |  |
| reasoning |  | — | Reasoning effort; reasoning tokens count toward max_tokens. |

## What can you do with Qwen3.5-Flash in Devin?

Turn a large collection of project code and documentation into focused explanations, draft patches, and review notes with Qwen3.5-Flash. Give each request a concrete question and an output format you can check against the project.

### Compare lengthy reports

Ask Qwen3.5-Flash to compare reports by topic, identify conflicting statements, and attach a source section to each finding.

### Extract records from messages

Turn customer messages into JSON records with an issue summary, requested action, and supporting text. Validate the resulting keys and types.

### Draft support replies

Provide the customer's question and the relevant policy text, then request a concise reply with unresolved questions listed separately.

### Explain unfamiliar code

Supply related code and ask for the execution path, dependencies, and a short list of places to inspect when a reported failure occurs.

### Prepare multilingual drafts

Translate product descriptions or support answers with a supplied glossary. Request a separate list of ambiguous phrases for review.

### Read dashboard screenshots

Extract visible labels, values, and units from a screenshot into a table. Ask it to mark unreadable values rather than fill gaps.

### Write recurring digests

Summarize newly supplied updates into decisions, deadlines, and open questions, retaining the source identifier for each item.

## Overview

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

## Qwen3.5-Flash in other agents

- [Amp](https://spiral.ai/integrations/amp/qwen3-5-flash)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/qwen3-5-flash)
- [Claude](https://spiral.ai/integrations/claude-desktop/qwen3-5-flash)
- [Claude Code](https://spiral.ai/integrations/claude-code/qwen3-5-flash)
- [Cline](https://spiral.ai/integrations/cline/qwen3-5-flash)
- [Codex](https://spiral.ai/integrations/codex/qwen3-5-flash)
- [Cursor](https://spiral.ai/integrations/cursor/qwen3-5-flash)
- [dots](https://spiral.ai/integrations/dots/qwen3-5-flash)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/qwen3-5-flash)
- [Gemini](https://spiral.ai/integrations/gemini/qwen3-5-flash)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/qwen3-5-flash)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/qwen3-5-flash)
- [goose](https://spiral.ai/integrations/goose/qwen3-5-flash)
- [Grok](https://spiral.ai/integrations/grok/qwen3-5-flash)
- [Grok Build](https://spiral.ai/integrations/grok-build/qwen3-5-flash)
- [Hermes](https://spiral.ai/integrations/hermes/qwen3-5-flash)
- [Instinct](https://spiral.ai/integrations/instinct/qwen3-5-flash)
- [Manus](https://spiral.ai/integrations/manus/qwen3-5-flash)
- [Muse](https://spiral.ai/integrations/muse/qwen3-5-flash)
- [Muse Code](https://spiral.ai/integrations/muse-code/qwen3-5-flash)
- [n8n](https://spiral.ai/integrations/n8n/qwen3-5-flash)
- [Notion](https://spiral.ai/integrations/notion/qwen3-5-flash)
- [OpenClaw](https://spiral.ai/integrations/openclaw/qwen3-5-flash)
- [opencode](https://spiral.ai/integrations/opencode/qwen3-5-flash)
- [Perplexity](https://spiral.ai/integrations/perplexity/qwen3-5-flash)
- [Raycast](https://spiral.ai/integrations/raycast/qwen3-5-flash)
- [VS Code](https://spiral.ai/integrations/vscode/qwen3-5-flash)
- [Windsurf](https://spiral.ai/integrations/windsurf/qwen3-5-flash)
- [Zapier](https://spiral.ai/integrations/zapier/qwen3-5-flash)

## More Devin guides

- [Qwen3.5-9B](https://spiral.ai/integrations/devin/qwen3-5-9b)
- [Qwen3.6 27B](https://spiral.ai/integrations/devin/qwen3-6-27b)
- [Qwen3.5-35B-A3B](https://spiral.ai/integrations/devin/qwen3-5-35b-a3b)
- [Qwen3.6 35B A3B](https://spiral.ai/integrations/devin/qwen3-6-35b-a3b)
- [Qwen3.5-27B](https://spiral.ai/integrations/devin/qwen3-5-27b)
- [Qwen3.6 Flash](https://spiral.ai/integrations/devin/qwen3-6-flash)

## Questions about Qwen3.5-Flash in Devin

**Can Devin use Qwen3.5-Flash?**

Yes. Add Spiral to Devin as an MCP server, and Devin can find and run Qwen3.5-Flash like any built-in tool.

**How much does Qwen3.5-Flash cost in Devin?**

Up to $0.001389 per run at the default inputs. $0.0845 per 1M input tokens, $0.338 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Devin plan is separate.

**Do I need an OpenRouter account to use Qwen3.5-Flash in Devin?**

No. Spiral runs Qwen3.5-Flash on its own OpenRouter account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Qwen3.5-Flash in Devin?**

About two minutes: add a custom MCP, then sign in to Spiral and approve, then check it.

**Does Devin wait while Qwen3.5-Flash runs?**

Qwen3.5-Flash runs in the background. Devin starts the run, checks on it and gives you the text, json when it's ready.

**What is Qwen3.5-Flash useful for?**

Use it to summarize long source material, extract records, draft support replies, and explain or generate code. Its 1 million token context window is useful when a task needs many documents considered together.

**How much can Qwen3.5-Flash read and write?**

It has a 1,000,000-token context window and supports up to 65,536 completion tokens. The context limit and output limit are different, so a large input capacity does not mean it can write an equally large response.

**When should I use thinking mode?**

Thinking is enabled by default in Alibaba's documentation. Try it for tasks that need several stages of analysis, such as diagnosing a bug or comparing conflicting requirements; use direct responses for straightforward extraction and rewriting.

**Can Qwen3.5-Flash return dependable JSON?**

Alibaba documents JSON Object structured output in non-thinking mode. That ensures valid JSON syntax, but does not guarantee the keys and types you need, so validate records before passing them downstream.

**Which languages does Qwen3.5-Flash support?**

Qwen reports coverage of 201 languages and dialects for Qwen3.5. For translation work, specify the target language, intended audience, and terms that must stay unchanged, then review a sample before processing a large collection.

