Browser 106

no file limit

Inspector ai/audio-transcription

Audio Transcriber

Overview

This tool runs OpenAI's Whisper speech-to-text model directly in your browser using WebAssembly, so your audio files are never uploaded to any server. Drop in MP3, WAV, M4A, OGG, FLAC, WEBM, or MP4 files and get accurate transcripts with optional word-level timestamps — no account required and nothing stored remotely. The Whisper model downloads once and is cached in your browser for offline use on future visits. Batch processing lets you queue multiple files and have them transcribed automatically, one after another.

Journalists and researchersTranscribe recorded interviews or field audio privately without sending sensitive source material to a third-party cloud service.
Podcast creatorsGenerate text transcripts from episode recordings to publish show notes, improve accessibility, or repurpose content.
Students and academicsConvert lecture recordings or dictated notes into searchable, editable text without a subscription service.
Multilingual transcriptionUse Whisper's automatic language detection or manually select a language to transcribe audio in dozens of languages.
Meeting notesTranscribe recorded calls or meetings locally without sharing confidential business audio with external APIs.

Key features

Local Whisper AI transcription
Multiple file batch processing
Model download with live progress
Per-file transcription progress
TXT download per transcript
Language auto-detect or manual selection
Optional word-level timestamps
Sequential processing for memory efficiency
Complete client-side privacy

Set

AI-hulpmiddelen 03

Neighbours

01 AI in de browser draaien 02 AI-afbeeldingen genereren 03 Factuur Generator 04 Markdown-editor met livevoorbeeld 05 Spreadsheet Converter 06 Tekstanalyse Tool

Specification

Command: ai.audio_transcription
Set: AI
Input: text
Output: —
Max. bestand: no limit
Batch: single file

Runtime

Engine: Transformers.js 3.x
Threads: — workers
Isolated: —
WebGPU: —
Memory: —

Transport 0 B

This set is the exception. the model is fetched once, then cached — your input never is — the traffic is stated here rather than hidden.

Keys

Zoeken: ⌘K
Panels: ⌘1 · ⌘2
Full width: F
All keys: ⌘/

How it works

1. Upload Audio

Drop one or more audio files, or use the file picker. Supports MP3, WAV, M4A, OGG, FLAC, WEBM, and MP4.

2. Choose a Model

Select a Whisper model. Tiny is fastest; Small is more accurate. The model downloads once and is cached in your browser for future use.

3. Transcribe

Click Transcribe Audio. Whisper runs entirely in your browser using WebAssembly. Your audio never leaves your device.

Frequently asked

Is my audio sent to a server?

No. All transcription runs locally using WebAssembly. Your audio never leaves your device.

Does the model download every time?

No. Once downloaded, the Whisper model is cached in your browser and reused automatically on future visits.

How large are the models?

Tiny is ~40 MB, Base is ~74 MB, Small is ~244 MB. Larger models are more accurate but slower and require more memory.

Can I transcribe multiple files?

Yes. Add multiple files to the queue and they will be processed one after another automatically.

What audio formats are supported?

MP3, WAV, M4A, OGG, FLAC, WEBM, and MP4. All formats are automatically converted to the 16kHz format Whisper requires.

What browsers are supported?

Chrome, Edge, Firefox, and Safari with WebAssembly support. Chrome and Edge offer the best performance with WebGPU acceleration.

Can I transcribe in languages other than English?

Yes. You can select a specific language or use Auto-detect to let Whisper identify the language automatically.

Keys

Zoeken: ⌘K
Panels: ⌘1 · ⌘2
Full width: F
All keys: ⌘/

Gerelateerd · AI-hulpmiddelen