Whisper
OpenAI's open-source speech recognition model for multilingual transcription and translation.
Commenters like it for running speech-to-text locally with accurate results1234, especially for dictation and multilingual transcription.5678 The most common complaint is transcription mistakes and hallucinations9101112, and some now prefer Parakeet for speed.131415
Why people recommend it
Runs locally · 3
Commenters run it on their own machines, phones and home servers without sending data to the cloud.
“a locally hosted whisper instance backed by Qwen 3.6”
“I use a local instantiation of whisper to transcribe it when I get home.”
Show 1 more
“Voice is on-device Whisper.”
Accuracy · 3
Several commenters say its transcriptions are accurate, often better than built-in OS speech-to-text.
“I much prefer whisper after playing with both. Maybe I'm not configuring Parakeet correctly, But the transcription that comes out of Whisper is usually pretty much spot on.”
“Whisper models in general seem to do a lot better than most built-into-OS/app models.”
Show 1 more
“Whisper remains the gold standard - you can run it locally with good accuracy across languages.”
Dictation · 3
People use it to dictate much of their typing, often through hotkey apps on the Mac.
“My favourite use is Whisper. I voice->text almost all of my typing now.”
“I’m using whisper with superwhisper on my mac.”
Show 1 more
“Current version sticks to FFmpeg + Whisper CLI locally, optimized for short chunks so you usually see text within about 1–2 seconds.”
Multilingual · 3
Valued for handling many languages, including transcribing foreign videos for translation.
“Whisper is still pretty good over all. At least if you need multilingual support.”
“I use Whisper to transcribe foreign language videos and then run them through a couple of models to translate into English. It works really well”
Show 1 more
“Whisper remains the gold standard - you can run it locally with good accuracy across languages.”
Video captions · 2
Creators use it to generate captions and subtitles in minutes instead of hours.
“Whisper has radically improved CC on my video content.”
“Whisper often makes mistakes which I now solve with an extra layer (qwen).”
Building block · 3
Seen as a small, specialized model that is easy to stack into larger pipelines.
“smaller specialised things like whisper, CLIP, SAM, etc.”
“it would make more sense to go straight to Whisper.”
Show 1 more
“audiobook -> text (via whisper)”
Caveats and dissent
Transcription errors · 3
Some report frequent mistakes or hallucinations that need correcting or a second model pass.
“If I switch to whisper it makes lots of mistakes”
“Whisper was catastrophically bad.”
Show 1 more
“Whisper often makes mistakes which I now solve with an extra layer (qwen).”
Parakeet faster · 3
Several commenters prefer Parakeet for speed, trading a little accuracy.
“I would choose Parakeet over Whisper in almost all situations these days.”
“accuracy slightly less than slower Whisper models”
Show 1 more
“Whisper is still pretty good over all. At least if you need multilingual support.”
Multiple speakers · 2
Works well for one speaker, but more speakers or complex repetitions produce repeated or poor output.
“Whisper is really good for 1 person speaking.”
“the audio files have some complex repetitions; the output was miserable.”
Heavy to run · 1
One user found it too heavy to run locally at good quality.
“I find Whisper too heavy to run locally with good quality”
Often compared with
1 moved from Parakeet to Whisper
1 moved from Whisper to ParakeetHandy7 comments mention both
1 moved from Whisper to Voxtral
Who recommends it in their own words
From the makers not counted in the ranking
“We use Whisper to transcribe the audio to text, and then use error-align[1] to align on the text.”
Threads where it came up
- Ask HN: What's the current best local/open speech-to-speech setup? 7 mentions
- Apple's new SpeechAnalyzer API, benchmarked against Whisper and its predecessor 4 mentions
- IKEA – 3D Guide 1 mention
- Ask HN: How do you turn videos into searchable text without uploading them? 1 mention
- Whistle: Speech to Text in 16.9 MB 1 mention
- Ask HN: Who wants to be hired? (October 2026) 1 mention
Data: whisper.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 42
“I use Whisper to transcribe foreign language videos and then run them through a couple of models to translate into English. It works really well”
“load whisper there and use the cloud GPUs to process them”
“When I use Whisper in the terminal, I can see it going back and correcting itself.”
“It used Whisper STT, Azure Neural TTS containers, SSML/prosody control”
“Whisper is still pretty good over all. At least if you need multilingual support.”
“Kimmi with whisper.”
“It used Whisper STT, Azure Neural TTS containers”
“Whisper often makes mistakes which I now solve with an extra layer (qwen).”
“I used to use Ollama and Whisper”
“It’s built with Tauri, a Rust backend, React/TypeScript, SQLite, GStreamer, Whisper, OCR, and VLM processing.”
“audiobook -> text (via whisper)”
“I use Gemma, Whisper and Qwen”
“a locally hosted whisper instance backed by Qwen 3.6”
“I find Whisper too heavy to run locally with good quality”
“If I switch to whisper it makes lots of mistakes”
“Whisper was catastrophically bad.”
“I would choose Parakeet over Whisper in almost all situations these days.”
“The whisper setup is something similar to what we have been using.”
“I use a local instantiation of whisper to transcribe it when I get home.”
“I actually built out a fine-tuning whisper pipeline with all local (smaller) models”
“smaller specialised things like whisper, CLIP, SAM, etc.”
“I made it a local Electron app with Whisper, ffmpeg”
“so I channel it all via Whisper.”
“Whisper models in general seem to do a lot better than most built-into-OS/app models.”
“Uses openai whisper for speech recognition”
“Voice is on-device Whisper.”
“Whisper remains the gold standard - you can run it locally with good accuracy across languages.”
“STT via Whisper”
“Whisper for STT”
“accuracy slightly less than slower Whisper models”
“Whisper is really good for 1 person speaking.”
“I’m using whisper with superwhisper on my mac.”
“it would make more sense to go straight to Whisper.”
“I'll probably go with whisper (with whisper.cpp) but has anyone compared it to vosk models?”
“streaming Speech-to-Text (using Whisper)”
“those are amazing models”
“My favourite use is Whisper. I voice->text almost all of my typing now.”
“Whisper works great, even the medium model is pretty good.”
“I use Prompt, Ever Terminal, Whisper, EC2 and Claude Code.”
“I also run a Whisper server and have built a voice-keyboard program.”
“I much prefer whisper after playing with both. Maybe I'm not configuring Parakeet correctly, But the transcription that comes out of Whisper is usually pretty much spot on.”
“Whisper has radically improved CC on my video content.”