Privro AI: a free, web-capable Whisper (Transcription) + captions + human like ai voice TTS web app

[avnshsingh]

Privro AI: a free, web-capable Whisper (Transcription) + captions + human like ai voice TTS web app

I built Privro AI, a free set of browser tools for transcription, captions, and AI voiceovers. The local tools run on your device, so your media stays in your browser.

Most transcription and video tools ask you to upload your files before you can do anything with them. Privro uses WebGPU and WebAssembly to process local projects on your machine. You do not need an account, API key, or subscription for local use.

Whisper transcription

Local Free Transcription Tool Turn video, audio, microphone input, or a URL into text with Whisper running on your device. WebGPU handles acceleration when available, with WASM as a fallback. Trascription by file format is also available such as convert mp4 to text.

You can export transcripts as TXT, JSON, CSV, SRT, VTT, XLSX, DOCX, or PDF.

Caption Studio

Caption Studio lets you style subtitles, preview them on your video, and export the finished video on your machine.

• Customize fonts, colors, placement, and karaoke effects
• Choose aspect ratios for YouTube, TikTok, Instagram, Reddit, X, and LinkedIn
• Export captioned video up to 4K

AI Voice

AI Voice turns text into a voiceover with on-device text-to-speech.

• Choose from natural AI voices in multiple languages
• Adjust voice settings for your project
• Generate audio without an API key or media upload

The local tools work on desktop, phone, and tablet. Once the required models are cached, you can use them offline.

Under the hood

Quantized Whisper models run through ONNX and Transformers.js with WebGPU or WASM. Text-to-speech models, subtitle rendering, and video export also run on the client. There is no backend media pipeline for these local tools.

Try it

No account required.

Feedback I want

If you try it, tell me:

1. How transcription speed feels on your GPU or CPU setup
2. Which Whisper models and AI voices work best for your language
3. Which Caption Studio controls or export options you still need

Thanks for taking a look.

[rekalit888556-code]

That's some amazing work by you avinash. I really like the work here specially related to your other tools capation studio and ai voice voice apart from your transcription at Privro AI. Gonna use daily for my daily day to day task.

[avnshsingh]

Thanks for feedback, i will work on it more and build something really useful for the community

[clionpar]

Running Whisper and TTS client-side via WebGPU is awesome. Not having to upload media to a server for quick captions is a huge win for both privacy and speed. Bookmarking this to test out offline model caching!

[avnshsingh]

Yeah, works great for transcription and captions. All local and privacy focus that's the main part

[anjali888555s]

I really the like work for transcription here, i use it regularly for day to day work. Other option are download whisper modal on to your computer not something easily accessible like website that this is really amazing. Thank you so much from the bottom of my heart and specially making it unlimited free forever ( I hope you will always have it this way) without no signup and that too privacy friendly.

[avnshsingh]

Hi Anjali! Thank you so much for your kind words appreciation like these help me in working harder and keep building such awesome tools. If you really like privro's online free local transcription service, there's good news for you that I will be making the web transcription part open source in upcoming few months for the community. This way people will be able to adopt it more open heartly.