mudler/parakeet.cpp

• highreadme#1

  Clarify README's opening to explicitly state Automatic Speech Recognition (ASR)

  原因：

  当前

  # parakeet.cpp
  
  **Brought to you by the LocalAI team**, the folks behind LocalAI, the open-source AI engine that runs any model (LLMs, vision, voice, image, video) on any hardware, no GPU required.
  
  [](https://huggingface.co/mudler/parakeet-cpp-gguf)
  [](LICENSE)
  [](https://github.com/mudler/LocalAI)
  
  parakeet.cpp is a C++17 inference port of NVIDIA's NeMo Parakeet speech-recognition models, built on ggml.

  复制粘贴的修复

  复制

  # parakeet.cpp: Fast C++ Automatic Speech Recognition (ASR) with ggml
  
  **Brought to you by the LocalAI team**, the folks behind LocalAI, the open-source AI engine that runs any model (LLMs, vision, voice, image, video) on any hardware, no GPU required.
  
  [](https://huggingface.co/mudler/parakeet-cpp-gguf)
  [](LICENSE)
  [](https://github.com/mudler/LocalAI)
  
  parakeet.cpp is a C++17 inference port of NVIDIA's NeMo Parakeet **Automatic Speech Recognition (ASR)** models, built on ggml.

  待办完成忽略
• hightopics#2

  Add comprehensive topics to improve categorization

  原因：

  当前

  (none)

  复制粘贴的修复

  复制

  asr, speech-recognition, c-plus-plus, ggml, nemo, parakeet, real-time, streaming, multilingual, cpu-inference, low-latency, speech-to-text

  待办完成忽略
• mediumreadme#3

  Add a 'Key Features' bulleted list to the README

  原因：

  复制粘贴的修复

  复制

  ## Key Features
  
  *   **C++17 Inference:** A native C++ port of NVIDIA's NeMo Parakeet ASR models, built on ggml.
  *   **Fast & Dependency-Light:** Optimized for CPU (and GPU via ggml backends) with no Python runtime needed at inference.
  *   **Comprehensive Model Support:** Covers all offline Parakeet families (CTC, RNNT, TDT, hybrid TDT-CTC, 0.6B/1.1B/110M sizes, English + multilingual v3).
  *   **Cache-Aware Streaming:** Includes end-of-utterance (EOU) detection for `parakeet_realtime_eou_120m-v1`.
  *   **Multilingual & Prompt-Conditioned:** Supports `nvidia/nemotron-3.5-asr-streaming-0.6b` (40+ locales) with language prompting.
  *   **Performance:** Faster than NeMo's PyTorch runtime on both CPU and GPU, with byte-identical transcripts.

  待办完成忽略

本次扫描解析到的品类 GEO 通道：google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• 品类问题

  How to achieve fast, dependency-light automatic speech recognition inference in C++ without Python?

  你：未被推荐
  AI 推荐顺序：

  1. Vosk API
  2. Picovoice Rhino Speech-to-Intent
  3. Porcupine Wake Word
  4. ONNX Runtime
  5. OpenAI's Whisper
  6. Conformer
  7. Mozilla DeepSpeech
  8. Kaldi
  9. TensorFlow Lite

  AI 推荐了 9 个替代方案，却始终没点名 mudler/parakeet.cpp。这就是要补上的差距。

  查看 AI 完整回答
• 品类问题

  What C++ libraries offer streaming multilingual speech recognition with end-of-utterance detection?

  你：未被推荐
  AI 推荐顺序：

  1. Vosk API (alphacep/vosk-api)
  2. Picovoice Rhino Speech-to-Intent (Picovoice/rhino)
  3. Picovoice Porcupine Wake Word (Picovoice/porcupine)
  4. Mozilla DeepSpeech (mozilla/DeepSpeech)
  5. WebRTC VAD (wiseman/webrtc-audio-processing)
  6. Google Cloud Speech-to-Text API (googleapis/google-cloud-cpp)
  7. Microsoft Azure Speech SDK
  8. Amazon Transcribe (aws/aws-sdk-cpp)

  AI 推荐了 8 个替代方案，却始终没点名 mudler/parakeet.cpp。这就是要补上的差距。

  查看 AI 完整回答

• Metadata completeness

  warn

  建议：

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of mudler/parakeet.cpp?

  pass
  ⚠AI 未点名 mudler/parakeet.cpp —— 很可能在说另一个项目

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• If a team adopts mudler/parakeet.cpp in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI 明确点名了 mudler/parakeet.cpp

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• In one sentence, what problem does the repo mudler/parakeet.cpp solve, and who is the primary audience?

  pass
  ⚠AI 未点名 mudler/parakeet.cpp —— 很可能在说另一个项目

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？