OfflineTTS puts private TTS and STT in your browser

OfflineTTS packages several audio workflows into a single browser-based toolkit: text to speech, speech to text, subtitle generation, and EPUB, PDF, or TXT to audio. The pitch is straightforward: it is free to use, requires no signup, and keeps core processing local in the browser.

The platform supports 4 local TTS engines and Whisper for transcription. According to the site, English TTS and Whisper STT can work offline after model download. It also offers TXT, SRT, and VTT exports, plus a workflow directory aimed at creators, accessibility use cases, study, and privacy-sensitive research.

The available engines and capabilities include:

• Kokoro: 54 voices across 9 languages
• Kitten: 8 expressions, designed as a lightweight local TTS option
• Piper: 25 voices, positioned as CPU-friendly
• Supertonic: 5 languages — English, Spanish, Portuguese, French, and Korean
• Whisper: 99 languages for transcription

OfflineTTS says all core tools run in the browser using WebGPU or WebAssembly. Beyond basic TTS, the service also handles audio transcription, timestamps, SRT subtitles, VTT caption exports, and document listening workflows.

The site frames this as an alternative to upload-first services such as ElevenLabs, NaturalReader, and Murf, emphasizing free usage, unlimited use, offline mode, and a browser-first privacy model rather than server-side processing. It also says there is no API key gate.

Recommended reading

Chrome’s Gemini tools reach UK desktop users

For exports, voice workflows support WAV and MP3, while transcription supports TXT, SRT, and VTT. The one caveat on commercial use: the site says generated speech is usable in most creator workflows, but users should check the upstream model terms before large-scale commercial deployment.