Free · No sign-up · Open source

AI Subtitles that run locally in your browser

Drop a video. Create AI subtitles on your device. Edit every word, make it your own, and export.

Launch Subtitle Studio Benchmarks & Code

100% localVideo stays on device

Works offlineOnce models are ready

Powered by AIWhisper speech models

Style each subtitle and see the result over your video.

Runs on your device

WebGPUGPU inference when available
OpenAI Whisper v3word-level timestamps
WebAssembly SIMDCPU fallback path
Transformers.jsclient-side inference
MediaBunnyin-browser demuxing

Drop your video here to begin

MP4, MOV, WebM — processed locally in your browser

or drag & drop

No cloud uploadRuns on your device

Features

Everything you need,
nothing you don’t.

Powerful AI features, a clean interface, and less effort.
Create better subtitles, faster.

AI Subtitle Generation

Whisper AI turns speech into text, directly in your browser. No cloud processing.

WebGPU accelerated

Smart Face Tracking

Subtitles follow the speaker. Face detection keeps text clear of the face automatically.

Split Subtitle Mode

Place subtitles above and below the speaker, or left and right, with face tracking.

Background Removal

Remove or blur the background with AI. Your video stays on your device.

25+ Fonts & Custom Styles

Choose colors, shadows, emphasis effects, and word-by-word highlighting.

100+ Languages

Whisper supports speech in over 100 languages, ready to select in the editor.

Export with Baked-in Subs

Download your video with subtitles as part of the picture, or save an SRT file.

Works Offline

Download the app and models once. Then use the cached tools without a connection.

Free & Built for Speed

No account or payment needed. Use your device’s GPU for fast, local processing.

Processing locallyAI runs in this browser tab

100% local

Audio extraction100%

Whisper transcription100%

Subtitle rendering100%

Your video stays on your device.Complete!

video.mp442 seconds · 1080pProcessed locally

Subtitles ready!Made on your device.

Privacy architecture

Your video stays
on your device.

The AI models download once. Then your browser does the work with WebGPU and WebAssembly. Your video files stay on your machine.

No video uploads

Your video is processed here, on your own device.

Private by design

Speech recognition runs in your browser, without sending your audio away.

Your device does the work

WebGPU and WebAssembly run AI on your own hardware.

You control your files

Edit in this session, then download the result when you are ready.

app/worker.tsWebGPU + WASM

import { pipeline } from "@huggingface/transformers";

// The `_timestamped` repos expose the alignment head the editor needs
// for word-level subtitle timing.
const transcriber = await pipeline(
  "automatic-speech-recognition",
  "onnx-community/whisper-base_timestamped",
  {
    device: "webgpu",
    dtype: { encoder_model: "fp32", decoder_model_merged: "q4" },
  },
);

const { text, chunks } = await transcriber(audio, {
  return_timestamps: "word",
});

Audio sent to a server: None
Compute backend: WebGPU, else WASM
Model weights: Cached on device

Engineered Architecture

Your browser is the entire production studio.

Traditional subtitling platforms upload heavy gigabytes of footage to remote cloud servers. FatahTech shifts the neural compute load straight to your client silicon.

Zero Video Uploads
Your confidential raw video files never leave local browser memory. There is no upload step and no server-side queue to wait in.
Private by Architecture
Inference runs on open-source Transformers.js and Whisper models inside this tab. No account, no tracking, and MIT-licensed code you can inspect.
Apple Silicon & NVIDIA Direct Acceleration
Hardware-accelerated matrix math through WebGPU compute shaders on Metal, DirectX 12, or Vulkan, with an automatic WebAssembly fallback.

How it works

Three steps. That’s it.

From your first word to your final video, all in one place.

Drop your video

Drag a video file into the browser. MP4, MOV, or WebM. Your video stays on your device.

Generate subtitles

Whisper AI transcribes your audio locally. Edit words, fonts, colors, position, and effects.

Export your video

Download your video with subtitles baked in. Share anywhere. The subs are part of the video.

Comparison

FatahTech vs Traditional Subtitle Tools

A capability comparison against the usual alternatives. No speed numbers: we have not published benchmarks, so we do not quote any.

Capability comparison between local browser-based transcription, cloud subtitle services, and a local command line tool.
Capability	FatahTech Subtitles	Typical cloud service	Local CLI (Whisper)
Where the video is processed	In your browser tab	Uploaded to a remote server	On your machine
Account required	No	Usually yes	No
Cost	Free, MIT licensed	Paid subscription	Free
Install required	No	No	Python, FFmpeg, a terminal
Works offline once cached	Yes	Needs a connection	Yes
Face-aware subtitle placement	Built in	Usually manual	Not available
Styled / word-level editing	Yes	Varies	Text output only

No sign-up · No API keys · 100% free

Ready to transcribe without limits?

Experience enterprise-grade AI subtitling directly on your computer hardware. Your video stays private, your speed stays instant.

Start subtitling now View on GitHub