# Qwen3 Coder Flash

Sparse MoE coding model with 30.5B total params, 3.3B active. Extended 1M context for repository-scale operations.

## Overview

- **Provider:** Qwen
- **Model ID:** `qwen3-coder-flash`
- **Type:** Chat
- **Canonical URL:** https://modelstack.cc/models/chat/qwen3-coder-flash
- **Version:** Coder Flash
- **Knowledge cutoff:** 2025-06
- **Best for:** Multi-file codebases, agentic coding workflows, complex refactoring, repository-wide analysis

## Pricing

| Usage | Price |
| --- | ---: |
| Input | $0.195 / 1M tokens |
| Output | $0.975 / 1M tokens |
| Cache read | $0.039 / 1M tokens |
| Cache write (5 minutes) | $0.24375 / 1M tokens |

## Specifications

- **Context window:** 1M
- **Maximum output:** 66K
- **Released:** 2025-09

## Capabilities

- chat

## API usage

Use this model ID in a ModelStack API request:

```json
{
  "model": "qwen3-coder-flash"
}
```

For the rendered page, visit https://modelstack.cc/models/chat/qwen3-coder-flash.