Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Turn PDFs and EPUBs into audiobooks; subtitles or videos into dubbed videos (including translation), and more. For free. Pandrator uses local models, including voice-cloning (instant, RVC-enhanced, XTTS fine-tuning) and LLM processing. It aspires to be a user-friendly app with a GUI, an installer and all-in-one packages.
| Date | Stars |
|---|---|
| 2026-07-24 | 593 |
| 2026-07-25 | 593 |
| 2026-07-28 | 596 |
| 2026-07-30 | 597 |
| 2026-08-06 | 597 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
45.0
growth rate 0.00%/day
<p align="center"> <img src="pandrator.png" alt="Pandrator" width="180" /> </p> # Pandrator Pandrator is a local-first workspace for creating audiobooks, subtitles, and voiceovers. It combines document preparation, transcription, optional LLM correction and translation, speech generation, detailed review, and export in one browser-based interface. The desktop installer and launcher manage Pandrator and its optional local model services. Docker and WSL are not required. The application itself is web-only. The retired Qt application is preserved on the `qt-maintenance` branch; Qt remains in `main` only for the desktop installer and launcher. ## TL;DR | If you want to… | Start with… | |---|---| | Create an audiobook with ready-made voices | **Kokoro** is the simplest lightweight starting point. Consult the [language table](#speech-generation-language-support) for alternatives. | | Clone a voice from a reference recording | Install **Qwen3 TTS Base** or another cloning model supporting your language. | | Create subtitles from audio or video | Install **CrispASR**. Use Whisper large-v3 for broad coverage or Parakeet TDT 0.6B v3 for its supported languages. | | Correct or translate text and subtitles | Configure a local or cloud LLM provider. This is optional for basic speech generation. | | Convert generated speech to another trained voice | Install **RVC**. It runs after speech generation and is optional. | Download the launcher from [GitHub Releases](https://github.com/lukaszliniewicz/Pandrator/releases). You can begin with only the components you need and add others later. Local models process content on your machine. Cloud LLM and speech providers are optional; when used, they may send content to an external service and incur charges. ## What Pandrator does ### Audiobooks - Imports plain text, pasted text, PDF, EPUB, DOCX, and MOBI sources. - Extracts structure and chapter markers, with OCR and a reviewable cleaning workflow for difficult PDF and EPUB files. - Includes a browser PDF editor with page stacks, left/right stacks, cropping, whiteouts, and deletion. - Applies deterministic text normalization and configurable segmentation before speech generation. - Optionally uses an LLM to clean a complete document or optimize small batches while generating. - Keeps generated speech as reviewable segments: edit, play as a playlist, mark, regenerate, compare takes, and select RVC variants. - Exports WAV, MP3, Opus, FLAC, or M4B with chapters, metadata, and cover art. ### Subtitles and voiceovers - Starts from SRT subtitles or common audio and video formats. - Transcribes media through CrispASR with word timestamps, VAD, and optional diarization controls. - Keeps transcription, correction, translation, subtitle composition, speech generation, synchronization, and export as separate, rerunnable steps. - Supports professional translation directly from an original transcript or from a corrected revision. - Provides side-by-side subtitle review with timing, text, split, and merge editing. - Creates subtitle-only exports or dubbed media with original, mixed, or dubbing-only audio and soft, burned, translated, original-language, or bilingual subtitles. ### Providers and voices - Connects to local TTS services, OpenAI, Google Gemini, and configurable custom speech endpoints. - Connects to local OpenAI-compatible LLM servers such as LM Studio as well as supported cloud providers. - Stores model-specific LLM temperature, reasoning, and cached/uncached token pricing defaults. - Manages recorded and uploaded reference samples, transcripts, and persistent previews of pre-built voices in the voice library. - Supports RVC model management and XTTS training as separate workflows. ## Installation ### Windows Download `PandratorInstaller.exe` from [Releases](https://github.com/lukaszliniewicz/Pandrator/releases) and run it. Choose an installation location and the local services you need. The same application is used later to launch
Excerpt of 18,575 characters
Read on GitHub
683
3
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:e73e38b8cd7e7a2d, topic:text-to-speech, topic:voice-cloning, desc:voice cloning