yan5xu/ququ

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

2271 stars · 230 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

QuQu is an open-source, free desktop voice dictation app built with Electron, designed as a privacy-first Wispr Flow alternative optimized for Chinese. It runs the FunASR Paraformer speech recognition model locally and uses a configurable LLM (OpenAI-compatible APIs, prioritizing Chinese models) to clean up and reformat transcribed text before pasting it at the cursor.

Use cases

• dictate text in Chinese instead of typing
• replace Wispr Flow without a subscription
• transcribe speech to text locally for privacy
• automatically remove filler words and fix speech errors with an LLM
• format dictated code terms like camelCase and snake_case
• reformat spoken notes into emails or code snippets with custom AI instructions

When to choose

• you need free, offline Chinese speech-to-text on macOS, Windows, or Linux
• voice data privacy matters and you want local ASR processing
• you want LLM-polished dictation output with support for Chinese AI models
• you want a global-hotkey dictation tool that pastes text into any app

When to avoid

• you need speech recognition for languages other than Chinese
• you cannot run a local Python FunASR service alongside the app
• you need a mobile or web-based dictation solution
• you require cloud-scale ASR with speaker diarization or batch transcription

Facets

application · maturity active

speech-recognition llm-inference nlp desktop-framework speech-processing developer-tools privacy large-language-models windows cross-platform python voice-dictation wispr-flow-alternative funasr local-processing chinese-speech-recognition ai-text-processing privacy-first speech-to-text natural-language-processing macos linux electron nodejs

Member repositories

For agents

markdown · JSON · MCP: product_card(name="yan5xu/ququ")