The most feature-complete open-source alternative to Google's NotebookLM — a self-hosted research platform where you upload PDFs, videos, audio files, and web pages into organized notebooks, then chat with your content, generate multi-speaker podcasts, and run semantic search across everything without sending a single byte to Google's servers. The podcast engine supports 1-4 fully customizable speakers with backstories, personalities, and expertise profiles, generating professional audio dialogue through OpenAI, ElevenLabs, Google TTS, or completely local text-to-speech via Kokoro for maximum privacy. Content processing uses token-based chunking with RAG-powered retrieval grounded in your uploaded sources, while both full-text keyword search and semantic vector search via SurrealDB enable conceptual discovery across all notebooks. The 18+ supported AI providers include OpenAI, Anthropic, Google Gemini, Groq, Ollama, LM Studio, and more — configurable per task so you can route cheap models to summarization and powerful models to analysis. Content transformations extract insights, generate summaries, create study guides, and produce structured outputs from any source material. The MCP integration connects Open Notebook to Claude Desktop, VS Code, and other MCP clients for seamless workflow integration. A full REST API on port 5055 enables complete automation of notebook management, source upload, and podcast generation. Deploy via Docker Compose with the application container, SurrealDB v2 on RocksDB, and optional TTS containers. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

OpenNotebookLM is a self-hosted research workspace that reads your sources with you, answering questions only from the material you gave it and showing the passage behind every claim. Import papers, spreadsheets, recordings, web pages, PDF links, YouTube videos or playlists, Markdown folders, Obsidian vaults, and Git repositories, then select which sources the chat may use as evidence. Stream answers with follow-up questions, hover any numbered citation to inspect the supporting passage and page, and let the app flag citations that look unsupported before you trust them. The Studio turns the same sources into cited reports in briefing, study guide, FAQ, or timeline formats, expandable mind maps you can pan and question, two-host audio overviews in deep dive, brief, critique, or debate styles, narrated slides exportable to PowerPoint, and tables whose cells carry their own citations. Generate flashcards and timed quizzes from what you read, mark cards for review, and resume saved progress later. Save cited answers and Studio outputs as notes, convert notes back into searchable sources, and export almost anything: conversations, projects, mind maps as SVG, flashcards as Anki text, audio as MP3. Share projects with viewer or editor roles, keep conversations private per person, and bring Claude, OpenAI, any OpenAI-compatible provider, or a local Ollama model, with extraction, OCR, Whisper transcription, and Kokoro audio synthesis running on your own server. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

With OpenReader, any EPUB, PDF, Word document, Markdown file, or plain text book transforms into an interactive reading workspace paired with natural text-to-speech playback and multi-chapter audiobook export. Documents render natively in their original page layout rather than flattening into plain text, using layout-aware block detection and cross-page stitching so multi-column academic papers, technical manuals, and textbooks flow in the right reading order while skipping headers and footnotes. As narration plays, word-level forced alignment highlights each spoken word directly on the page in real time, and progressive background generation starts audio playback on the first segment while synthesizing upcoming pages ahead of your cursor. Readers can connect local speech engines such as Kokoro, KittenTTS, and Orpheus or cloud voice providers, select language-aware voices for multilingual books, adjust playback speed, and switch among thirteen light and dark reading themes. Cached audio segments persist across seeks and browser reloads, allowing the background worker to assemble complete M4B or MP3 audiobooks for offline listening on mobile devices while syncing reading progress, queues, and library imports across sessions. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Folivox turns an ebook sitting on your server into a finished audiobook read aloud in a voice you choose, cloned from a short sample, or performed by a whole cast of characters. Drop in an FB2, EPUB, DOCX, PDF, or TXT file and the studio detects the book language, restores chapters, poems, and epigraphs, cleans page furniture away, and spells numbers, dates, money, and abbreviations out in full for nine fully supported languages. Narration runs locally through CosyVoice 3 or the multi-speaker Fish Audio Drama 3, which can give each character their own voice like a radio play, or through ElevenLabs, OpenAI, Gemini, or Cartesia with your own key after a cost estimate you approve. A markup editor finds the dialogue, assigns lines to speakers, and lets you fix mistakes and listen before committing to a ten-hour render. Every fragment is listened back to with Whisper and re-read until the best take survives, while a crash-safe journal means a stopped render resumes where it left off and edited chapters redo only what changed. The finished book arrives as one chaptered M4B with cover and metadata, or one MP3 per chapter, loudness-normalised and ready for any player. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Voicebox is a voice studio that clones a voice from a few seconds of audio and then speaks any script in it, all on hardware you control. Eight speech engines cover multilingual cloning, preset voices, voices described in plain words, and expressive tags like laughs and sighs, across 23 languages. Long scripts are split at sentence boundaries and crossfaded, so chapters and articles render in one pass. A post-processing rack adds pitch shift, reverb, delay, chorus, compression, and filters, with presets like Robotic and Radio. The Stories editor lays voices on a multi-track timeline for podcasts and dialogues, then exports the mix. Voice profiles can be recorded in the app, imported, exported, and given personas that a bundled local language model uses to compose, rewrite, or respond. Global dictation turns a held hotkey into transcribed text with optional cleanup of filler words. A built-in MCP server and REST API let coding agents call a single speak tool and talk back in a cloned voice. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

With 26,000 GitHub stars and consistent victories over ElevenLabs in blind evaluations, Chatterbox delivers state-of-the-art text-to-speech with zero-shot voice cloning requiring only 5 seconds of reference audio. The model family spans three architectures: Chatterbox Multilingual V3 (500M parameters, 23+ languages including Arabic, Chinese, Japanese, Korean, Hindi, French, German, Spanish, and Portuguese), Chatterbox-Turbo (350M parameters optimized for voice agents with a single-step distilled decoder achieving ~200ms time-to-first-speech), and Chatterbox-Nano (110M parameters running 3x faster than realtime on 8 CPU cores for edge deployment). Unique among open-source TTS systems, Chatterbox introduces emotion exaggeration control — adjusting intensity from monotone to dramatically expressive via a single parameter — and native paralinguistic tagging where tokens like [laugh], [cough], [chuckle], and [gasp] inject natural vocal reactions inline without post-processing. The alignment-informed inference pipeline eliminates hallucinations and repetition artifacts common in autoregressive TTS. Built-in PerTh neural watermarking embeds imperceptible forensic identifiers in generated audio for provenance tracking. Trained on 500,000 hours of cleaned speech data across all supported languages. Voice conversion scripts enable transforming existing audio into any cloned voice. Deploy via pip install with PyTorch, serve through Gradio interfaces or custom FastAPI endpoints, and expose via HTTP streaming or WebSocket for sub-200ms conversational applications. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Every morning, TLDR Radio converts the day's TLDR newsletters into chapterized podcast episodes you listen to instead of read, generated entirely on your own server. Pick a date at the Broadcast Desk, hit Broadcast, and a few minutes later chapterized episodes wait in your browser, one per newsletter with each story as its own track. All fourteen TLDR newsletters can sit on the Desk for manual runs, join the overnight build, or both, and the overnight selection shows an estimated listening time measured from your own episodes. Scheduled builds retry until the day's issue appears and can send an ntfy notification when the audio is ready. The player groups your library by day, shows each chapter with its section heading, source summary, and link to the original, and offers 15-second skips, playback speed up to 2.5x, auto-advance, resume, and keyboard shortcuts. Sponsor segments become their own clearly marked chapters in a different voice, announced out loud so an ad is never mistaken for news, and easy to skip. Sixty-eight voices can be auditioned in Settings, and every episode downloads as a single tagged MP3 with album, title, and author set for podcast and audiobook apps. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

OpenTranscribe is a self-hosted transcription platform that converts your audio and video library into searchable, speaker-labeled transcripts, then lets you summarize and chat with those recordings. Upload files up to 15 GB, record from the browser microphone, or paste a link from YouTube and a thousand other sites, and playlists queue up for batch processing automatically. WhisperX produces word-level timestamps in over 100 languages, while the diarization engine works out who spoke when, corrects turn boundaries, and recognizes the same voice again across different recordings. Speaker profiles accumulate a voice fingerprint, speaking-time analytics, and merged identities, so a regular guest is named correctly in every episode without relabeling. Point it at an LLM and recordings gain structured summaries with action items, suggested tags and collections, and a chat interface that answers questions about your transcripts with citations, grouped into projects that pin their source recordings and standing instructions. Leave the model unconfigured and transcription, diarization, full-text and semantic search, and redaction still run entirely on local models. Watch folders, S3 buckets, and network shares import new media on a schedule with three-layer deduplication and automatic stitching of split recordings, and everything exports as subtitles, Word or PDF documents with citeable numbered lines, or Markdown. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

Audio lectures become structured, Markdown-formatted notes in about a minute with ScribeWizard (also known as GroqNotes). Upload an MP3, WAV, or M4A file - or paste a YouTube link - and the app runs a three-stage pipeline on Groq's LPU inference hardware: Whisper Large v3 transcribes the audio, a larger Llama model drafts a comprehensive outline of the material, and a faster Llama model fills each section with detailed content. This scaffolded prompting strategy is the core idea: the strong model handles structure where quality matters most, the fast model handles volume, and Groq's 1200+ tokens-per-second inference keeps the whole process near real time. Output renders as clean Markdown with support for tables and code blocks, and finished notes download as text or PDF. Model selection is configurable - swap in other Groq-hosted open models like Mixtral or Gemma to trade speed against quality or work around rate limits. Built as a single Streamlit app by Benjamin Klieger at Groq, it needs only a Groq API key to run, making it one of the simplest self-hosted AI tools to operate.

Your lecture slides go in one end, and a personal tutor who has actually read them comes out the other: OpenTutor turns uploaded course material into structured notes, flashcards, quizzes, and a chat tutor that answers from your own documents. Drop in a PDF, DOCX, or PPTX file, or connect Canvas LMS, and within seconds the workspace fills with generated study material organized into twelve composable blocks you can rearrange as your habits change. Ask the tutor questions and every answer cites the source material it came from, with a Socratic mode that prods you toward the answer and depth that adjusts when your replies show fatigue or repeated mistakes. Practice with seven question types, from multiple choice and fill-in-blank to matching, ordering, and coding, and wrong answers are tracked with diagnostic feedback that reshapes later difficulty. Flashcards return on an FSRS 4.5 spaced-repetition schedule tuned to what you are about to forget, and an experimental knowledge graph maps concept mastery, prerequisites, and weak spots into a learning path. Plan study sessions on the calendar, watch progress in analytics, export your work, and connect ten or more LLM providers or a local Ollama model so your documents never have to leave your server. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Scholarsome is a self-hosted flashcard platform where you create study sets, share them with other learners, and drill them through flashcards and automatically built quizzes. Start a set from scratch in the card editor, writing terms and definitions with rich formatting, images, and LaTeX expressions for math and science notation. Bring your existing material with you instead: decks import from Anki, Quizlet, and CSV files, and finished sets export back to the same formats so nothing is trapped inside. Share a set with classmates or keep it private with visibility controls, then let everyone study from the same central copy with no device syncing to manage. Multiple learning modes mimic real flashcards for relaxed review, while the quiz builder generates fill in the blank, true or false, and multiple choice questions from your cards for a stricter test. Teachers can assemble a class deck once and publish it to every student at the moment it is ready. Because the whole system runs on your own server, study data stays under your control rather than behind a subscription paywall. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

Podcast enthusiasts and homelab archivers automate digital episode preservation and stream personal audio feeds using PodFetch, an open-source self-hosted podcast manager that combines scheduled RSS downloads, Podcasting 2.0 transcript indexing, and browser playback into a modern listening platform. Listeners can subscribe to international podcast channels through direct feed URLs, Apple Podcasts directory searches, or bulk OPML subscription imports, configuring background polling timers to download new releases automatically upon publication. The integrated audio player supports variable speed playback, episode bookmarks, listening histories, and fullscreen visualization modes directly inside responsive desktop and mobile browser sessions. Advanced search capabilities leverage Podcasting 2.0 synchronized transcripts and automated speech-to-text generation via OpenAI-compatible Whisper APIs, allowing collectors to run full-text keyword searches across every spoken phrase in their audio archive. Built-in GPodder API synchronization maintains subscription lists and playback positions across popular smartphone apps like AntennaPod and native companion clients without relying on commercial cloud accounts. Server operators can assign multi-user accounts with private favorites lists, export customized RSS aggregation streams, and route storage paths between fast solid-state drives and large secondary mechanical hard disks. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

PinePods delivers a centralized, self-hosted podcast ecosystem that liberates personal listening histories, audio files, and show subscriptions from commercial streaming walled gardens. Users can discover new shows through built-in search connectors across Podcast Index, iTunes, and YouTube, streaming audio and video episodes with variable playback speed, chapter markers, and synchronized transcripts. A persistent queue and custom playlist builder keep media sequences orderly across web browsers, desktop installations, and mobile clients with native CarPlay and Android Auto support. Built-in gPodder server synchronization enables seamless cross-device progress tracking with third-party podcast apps like AntennaPod. Automatic episode downloader engines monitor subscribed feeds, archiving new releases to local server storage according to customizable retention rules and storage quotas. Multi-user authentication secures household access through OIDC single sign-on and two-factor TOTP verification, while an integrated terminal client lets command-line enthusiasts manage queues directly over SSH. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. GPL-3.0 licensed.

Eliminating repetitive sponsor reads and dynamic mid-roll commercials before an episode even reaches your headphones, MinusPod operates as a self-hosted podcast ad-removal server and RSS feed manager. You subscribe to shows inside the web dashboard using PodcastIndex search or OPML import, and the server generates clean replacement RSS feeds compatible with Pocket Casts, Overcast, Apple Podcasts, and Audiobookshelf. When a new episode drops via scheduled polling or instant Podping notifications, the pipeline transcribes the audio locally with faster-whisper or Parakeet, or routes it through a remote transcription API. Your choice of LLM provider, including Ollama, Anthropic Claude, OpenRouter, or any OpenAI-compatible endpoint, scans overlapping transcript windows to pinpoint sponsor reads, cross-promotions, and intros. Programmatic audio analysis reinforces AI detection by measuring EBU R128 loudness jumps, Chromaprint acoustic fingerprints, show jingles, and dynamically inserted ad transitions before FFmpeg splices the audio or marks segments as skippable Podcasting 2.0 chapters. A browser-based Ad Editor lets you audition pre-cut and post-cut audio, nudge start and end timestamps, and confirm detections that train global, network, and show-specific sponsor patterns. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Insights LM gives you a NotebookLM-style research assistant that runs on your own infrastructure and answers strictly from the documents you feed it. Build a notebook for a project, drop in PDFs, pasted text, website links, and audio files, then ask questions in plain language and get answers grounded in those sources rather than the open internet. Every answer carries citations you can click to jump straight to the passage it came from, so verifying a claim takes seconds instead of a fresh search. Findings worth keeping become notes saved alongside the notebook, building a private knowledge base that grows with your research. When reading is too slow, generate an audio overview and listen to two hosts discuss your material like a podcast episode, perfect for absorbing a long report on a walk. Teams use it to interrogate internal documentation, onboard new hires against company knowledge, and brief themselves before meetings without uploading confidential files to a third-party service. Because the whole stack is yours, you choose which AI models do the reading, including local models for fully private work. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.