No signup·Free
AI Transcript Generator, Free & Private
No account, no cloud processing — an open speech-recognition model runs entirely in your browser.
MP3 · WAV · M4A · AAC · FLAC · OGG · MP4 · MOV · WebM
Supports MP3, WAV, M4A, AAC, FLAC, OGG, MP4, MOV, and WebM.
Your files never leave your device. Transcription runs locally in your browser.
How the AI works
- Whisper speech recognition
- The transcription model is OpenAI's open-source Whisper, running fully on-device — the same underlying technology used by many paid transcription services.
- WebGPU acceleration
- When your browser supports WebGPU, the model runs on your GPU for faster transcription. Without it, the app automatically falls back to a slower CPU-only mode — it still works either way.
- Local inference means local privacy
- Because the AI model runs in your browser, your audio and video never have to leave your device to be transcribed.
Local AI vs. cloud AI
Local (this tool)
Private and free, but limited by your device's speed and memory — best for shorter files.
Cloud (PodText)
Faster on long files, higher-accuracy models, and background processing — a good fit when a file is too large or slow to process locally.
Need better accuracy, summaries, or background processing? Continue with PodText. Continue with PodText
Frequently asked questions
How accurate is the AI?
Accuracy depends on the model and audio quality — the default Tiny model favors speed, while the larger Base model trades some speed for higher accuracy.
Does it need an internet connection?
Only to download the Whisper model the first time. After that, the model is cached in your browser and transcription itself doesn't require a network connection.
Is my audio used to train the AI?
No. Your files are never uploaded anywhere, so they can't be used for training.