All tasks / Voice & audio / Transcription
Best local AI for transcription
Open-weights models and tools that run entirely on your own hardware, ranked from public benchmarks. Local tools are rated by the best open model they run.
ATop ranked: Qwen3-ASR-1.7B (Alibaba (Qwen)), 5.4% on Hugging Face Open ASR Leaderboard (English, short-form average WER).33 options listed · data verified 19 Oct 2026
Ranked list
| Tier | Name | Approach | Runs | Rated by | Price | Skill |
|---|---|---|---|---|---|---|
| A |  Qwen3-ASR-1.7BAlibaba (Qwen) | Models | Local | 5.4% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~6 GB VRAM | — |
| A |  Voxtral Small (24B-2507)Mistral AI | Models | Local | 2.8% on Artificial Analysis Speech to Text: AA-WER v2 (non-streaming) | ~55 GB VRAM | — |
| A |  Higgs Audio v3 STTBoson AI | Models | Local | 5.9% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~6 GB VRAM | — |
| A |  Granite Speech 4.1 2BIBM | Models | Local | 6% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~6 GB VRAM | — |
| A |  MOSS-Transcribe-DiarizeOpenMOSS (Fudan University / MOSI) | Models | Local | 6.2% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~2 GB VRAM | — |
| A |  distil-large-v3.5Hugging Face | Models | Local | 6.8% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~2 GB VRAM | — |
| A |  Whisper large-v3OpenAI | Models | Local | 4.1% on Artificial Analysis Speech to Text: AA-WER v2 (non-streaming) | ~10 GB VRAM | — |
| A |  VoiceboxJamie Pine (open source) | Apps | Local | via Whisper large-v3 | Free | beginner |
| A |  whisper.cppggml-org (open source) | Engines & APIs | Local | via Whisper large-v3 | Free | advanced |
| A |  HandyCJ Pais (open source) | Apps | Local | via Whisper large-v3 | Free | beginner |
| A |  faster-whisperSYSTRAN (open source) | Engines & APIs | Local | via Whisper large-v3 | Free | advanced |
| A |  WhisperXMax Bain (open source) | Engines & APIs | Local | via Whisper large-v3 | Free | advanced |
| A |  BuzzChidi Williams (open source) | Apps | Local | via Whisper large-v3 | Free | beginner |
| A | sherpa-onnx k2-fsa (Next-gen Kaldi) | Engines & APIs | Local | via Whisper large-v3 | Free | advanced |
| A | Vibe thewh1teagle (open source) | Apps | Local | via Whisper large-v3 | Free | beginner |
| A |  VoiceInkPrakash Joshi Pax (open source) | Apps | Local | via Whisper large-v3 | — | beginner |
| A |  WhisperKitArgmax | Engines & APIs | Local | via Whisper large-v3 | Free | advanced |
| A |  WhisperLiveCollabora | Engines & APIs | Local | via Whisper large-v3 | Free | advanced |
| A |  Speachesspeaches-ai (open source) | Engines & APIs | Local | via Whisper large-v3 | Free | advanced |
| A | Whishper pluja (open source) | Apps | Local | via Whisper large-v3 | Free | advanced |
| A |  AikoSindre Sorhus | Apps | Local | via Whisper large-v3 | Free tier | beginner |
| A |  MacWhisperJordi Bruin (Good Snooze) | Apps | Local | via Whisper large-v3 | Free tier | beginner |
| A |  SpokenlySpokenly | Apps | Local | via Whisper large-v3 | Free | beginner |
| A |  SuperwhisperSuperwhisper | Apps | Local | via Whisper large-v3 | Free tier · $8.49/mo | beginner |
| B |  Canary-Qwen-2.5BNVIDIA | Models | Local | 4.3% on Artificial Analysis Speech to Text: AA-WER v2 (non-streaming) | ~6 GB VRAM | — |
| B |  Fun-ASR-Nano-2512Alibaba (FunAudioLLM) | Models | Local | 7.2% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~2 GB VRAM | — |
| B | Parakeet TDT 0.6B v3 NVIDIA | Models | Local | 4.5% on Artificial Analysis Speech to Text: AA-WER v2 (non-streaming) | ~2 GB VRAM | — |
| B |  cohere-transcribe-03-2026Cohere | Models | Local | 4.6% on Artificial Analysis Speech to Text: AA-WER v2 (non-streaming) | ~6 GB VRAM | — |
| B |  omniASR-LLM-7B-v2Meta (FAIR) | Models | Local | 8.1% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~16 GB VRAM | — |
| B |  VibeVoice-ASRMicrosoft | Models | Local | 8.2% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~19 GB VRAM | — |
| B | Voxtral Mini 4B Realtime (2602) Mistral AI | Models | Local | 8.4% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~10 GB VRAM | — |
| B |  NVIDIA NeMo SpeechNVIDIA | Engines & APIs | Local | via Canary-Qwen-2.5B | Free | advanced |
| C | Nemotron 3.5 ASR Streaming 0.6B NVIDIA | Models | Local | 9.8% on Hugging Face Open ASR Leaderboard (English, short-form average WER) | ~2 GB VRAM | — |
Tiers come only from public benchmarks (Hugging Face Open ASR Leaderboard (English, short-form average WER), Artificial Analysis Speech to Text: AA-WER v2 (non-streaming)); tools without a result of their own inherit the tier of the best model they offer. How we rank.