feat(asr): 支持 Qwen-Audio-3.0-ASR-Flash 整段转写

[MurphyLo]

User description

摘要

为阿里云百炼 ASR 增加 qwen-audio-3.0-asr-flash 模型支持。

该模型复用现有 DashScope multimodal-generation/generation 录音文件转写实现。

参考文档：
https://help.aliyun.com/en/model-studio/non-realtime-speech-recognition-user-guide

新增

• 在阿里云百炼模型列表中增加 qwen-audio-3.0-asr-flash。
• 增加该模型的路由和配置校验。
• 增加对应的请求体、模型路由及白名单测试。

测试计划

• INSTALL=0 ./scripts/build-mac.sh
• macOS ARM64 release 构建成功并生成 DMG。
• 在 OpenLess 中选择 qwen-audio-3.0-asr-flash，完成录音并确认转写结果正常插入。

---

PR Type

Enhancement, Tests

---

Description

• Add qwen-audio-3.0-asr-flash to DashScope batch ASR

• Unify routing, validation, and provider model lists

• Update frontend recorded-file detection for new model

• Extend tests for routing, validation, and request body

---

Diagram Walkthrough

flowchart LR
  A["Bailian model selected"] --> B{"Is qwen-audio-3.0-asr-flash or fun-asr-flash?"}
  B -- "Yes" --> C["DashScope multimodal ASR client"]
  B -- "No" --> D["Other Bailian protocol"]
  C --> E["Recorded-file batch transcription"]

Loading

File Walkthrough

Relevant files

Enhancement

dashscope_multimodal.rs Add qwen-audio model constant and validation                          openless-all/app/src-tauri/src/asr/dashscope_multimodal.rs Add QWEN_AUDIO_MODEL constant for the new model Introduce is_supported_model helper for both models Extend body-shape tests to cover the new model +23/-12  providers.rs Expose qwen-audio model in provider model lists                    openless-all/app/src-tauri/src/commands/providers.rs Add new model to the Bailian provider model list Add new model to DashScope multimodal static aliases Update connectivity sample comment to cover both models +7/-2      coordinator.rs Route and validate qwen-audio multimodal ASR model              openless-all/app/src-tauri/src/coordinator.rs Route qwen-audio-3.0-asr-flash to multimodal provider Use is_supported_model in routing and validation Update error messages with new supported models Add routing and validation tests +22/-5    ProvidersSection.tsx Recognize qwen-audio model as recorded-file ASR                    openless-all/app/src/pages/settings/ProvidersSection.tsx Recognize new model as recorded-file ASR in UI Add comment that vocabulary field is not yet wired +4/-2

---

[github-actions]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 PR contains tests

🔒 No security concerns identified

⚡ No major issues detected

[appergb]

已审查 Qwen-Audio 3.0 ASR 路由、模型校验、请求体与响应解析；阿里云官方文档确认该模型复用现有同步 multimodal-generation 协议。已在最新 beta 上与 #874/#875 联合验证：819 个 Rust 测试、26 个前端/契约测试及生产构建通过。