SurfSense

SurfSense: a self-hosted NotebookLM alternative I'm one of the maintainers, so this is self-promotion. I want honest feedback more than upvotes. What it is SurfSense is an open-source (Apache-2.0) NotebookLM alternative: cited Q&A over your documents, mind maps, flashcards, quizzes and podcasts, and an agent that writes Word, PDF, PowerPoint and Excel files and edits your own files as a new copy, with Word edits coming back as tracked changes. It started as a desktop app, and since 2.3.0 the same app runs as one container. Why self-host it NotebookLM sends every document to Google and only works with Gemini. That rules it out for contracts, client files, medical or company documents. SurfSense keeps your documents on your own server and works with any model. The container - One image, ghcr.io/modsetter/surfsense, with one service and one /data volume. No Postgres, Redis or other services to run alongside it. - Caddy is the only thing listening. Everything else stays on the container's loopback, and it needs no Docker socket. - Local models run inside it on the CPU (llama.cpp, plus local OCR, embeddings and voices), or you can point it at Ollama, vLLM or any OpenAI-compatible API on your network. - One password, no accounts. You can set the password on first visit or with SURFSENSE_PASSWORD, or turn sign-in off when your own reverse proxy handles it. - HTTPS on your LAN through Caddy's own certificate authority, or with Let's Encrypt for a public domain. Getting it running curl -fsSL raw.githubusercontent.com | bash Or grab docker/surfsense/docker-compose.yml and .env.example from the repo and run docker compose up -d. It's on port 4929, and 4943 for HTTPS. Running the script again updates it. Good to know - amd64 and CPU only for now. A GPU image and arm64 are planned. - The image is several GB: Docling, torch and about 1 GB of bundled models. Chat models download into /data at run time. - Docker's published ports bypass ufw, so if the server is reachable from the internet, set SURFSENSE_PASSWORD before the first start. How we used AI in development Most of the code is written with AI coding assistants, mainly Claude Code and some Cursor, with maintainers designing the features, reviewing every change and testing the releases. New behaviour starts with a failing test, and each feature has an architecture doc the assistants read before changing it. Repo: MODSetter/SurfSense Docs: surfsense.com

SurfSense: an open-source NotebookLM for Linux I'm one of the maintainers, so this is self-promotion. I want honest feedback more than upvotes. SurfSense is a free, open-source (Apache-2.0) desktop app. Think NotebookLM, but it runs on your machine, works with any model, and edits your files. There's no account, nothing reaches the network until you allow that host, and usage statistics are optional and off until you say yes. Linux details - x64 builds as AppImage, deb and rpm, about 1 GB each, with a SHA256SUMS file on every release. - Runs on Ubuntu 22.04 and RHEL 9 or newer. - Local models run on your GPU through Vulkan, which works with NVIDIA, AMD and Intel graphics. - Pick your model during setup: download a local one, connect Ollama, LM Studio, vLLM or any OpenAI-compatible API, or use your ChatGPT plan. What it does - Cited Q&A from your PDFs, Office files, CSV, Markdown, images and audio recordings. Scans are read with OCR, and recordings are transcribed locally. - Creates Word documents, PDFs, slides, spreadsheets and web pages with their own charts and interactive figures. It can start from your own .docx or .pptx as a template. - Edits your files as a new copy: tracked changes in Word, formulas in Excel, text on slides, and forms, merges and splits in PDFs. Your original is never touched. - Studio: mind maps, flashcards, quizzes and podcasts, like NotebookLM. - MCP plugins: Notion and DeepWiki from official third party list, or any server by URL. They stay off until you allow them. Where it's weaker: no video overviews, and the local voices don't sound like Google's. Download: surfsense.com Source: MODSetter/SurfSense

Open-source NotebookLM replacement that keeps your documents off Google's servers I'm one of the maintainers, so this is self-promotion. I want honest feedback more than upvotes. NotebookLM (now Gemini Notebook) is useful, but every document you add goes to Google, and you can only use Gemini. SurfSense is an open-source (Apache-2.0) desktop app for Windows, macOS (Apple Silicon) and Linux that does the same job on your own machine. Why it fits here - No account, and no Google sign-in. - Your files stay on your disk. Parsing, OCR, search and the model can all run locally. - Nothing reaches the network until you allow that host, and that includes app updates. - Usage statistics are optional and off until you say yes; they never include an ID or your content. - You choose the model: a local one it downloads for you, Ollama or LM Studio, or any OpenAI-compatible API. What it does - Cited answers from your PDFs, Office files, CSV, Markdown, images and audio recordings. - Mind maps, flashcards, quizzes and podcasts, voiced on your machine. - An agent that writes Word documents, PDFs, slides, spreadsheets and web pages, and edits your own files as a new copy. Word edits come back as tracked changes. Where it's worse than NotebookLM - No video overviews. - The local voices don't sound like Google's hosts. MODSetter/SurfSense

Update on SurfSense, the open-source NotebookLM alternative: now works with your Claude or ChatGPT plan I'm one of the maintainers, so this is self-promotion. I'm posting here because you know NotebookLM best, and I want honest feedback more than upvotes. SurfSense is an open-source (Apache-2.0) desktop app for Windows, macOS (Apple Silicon) and Linux. There's no account and no terminal. It downloads a local model for you. You can also connect any OpenAI-compatible API, or use your ChatGPT plan. It stays offline unless you choose to connect something, and that includes app updates. New since last time: - Use your Claude plan. If you pay for Claude Pro, Max or Team and have Claude Code installed, you can use that plan in both Q&A and Agent chats. - Faster scanned PDFs and photos. OCR goes first, and on a Mac it uses Apple's Vision framework. - Smaller downloads. Installers are smaller, and updates download only what changed. - Better web pages. The agent and Studio write web pages with their own charts, diagrams and interactive figures. Everything is built into the page, so it opens offline. - Audio overviews are MP3 now, and audio recordings can be sources. Two ways to chat: - Q&A: ask questions and get answers cited from your sources, like NotebookLM. It works with any model. - Agent: give it a task and it works through it step by step. It searches and reads your sources, then writes Word documents, PDFs, slides and spreadsheets. It can start from your own .docx or .pptx as a template. It needs a model that can call tools. Same as NotebookLM: mind maps, flashcards, quizzes, audio overviews (MP3, with a transcript), and editable .pptx, .docx and .xlsx exports. What NotebookLM can't do: - It edits your files and saves the result as a new copy. Your original is never touched. Word edits come back as tracked changes and comments. It updates Excel values and formulas, and replaces text on slides. - PDF tools: fill and flatten forms, add watermarks and page numbers, and merge, split, rotate or reorder pages. - It analyses CSV and Excel files and charts the results. - It reads scanned PDFs and images with OCR (Latin, Chinese and Japanese script). It also transcribes audio recordings on your machine, so you can search and cite them. - MCP plugins: the agent can use MCP servers from our plugin registry, such as Notion and DeepWiki, or any server you add by URL. They stay off until you allow them, and you choose which tools it can use and which ones ask first. - It works with any model. Where it's worse: - No video overviews. - Turning Word files or slides into PDFs needs LibreOffice. It's a separate download that you turn on in Settings. Your files stay on your disk. Usage statistics are optional and off until you say yes; they never include an ID or your content. The only limit on how many sources you add is your disk space. MODSetter/SurfSense

Open-source NotebookLM that edits your files A quick tour of SurfSense I'm one of the maintainers, so this is self promotion. I'm posting here because you know NotebookLM best, and I want honest feedback more than upvotes. SurfSense is an open-source (Apache-2.0) desktop app for Windows, macOS (Apple Silicon) and Linux. There's no account and no terminal. It downloads a local model for you. You can also connect any OpenAI-compatible API, or sign in with a ChatGPT Plus or Pro plan. It stays offline unless you choose to connect something, and that includes app updates. Two ways to chat: - Q&A: ask questions and get answers cited from your sources, like NotebookLM. It works with any model. - Agent: give it a task and it works through it step by step. It searches and reads your sources, then writes Word documents, PDFs, slides and spreadsheets. It can start from your own .docx or .pptx as a template. It needs a model that can call tools. Same as NotebookLM: mind maps, flashcards, quizzes, audio overviews (MP3, with a transcript), and editable .pptx, .docx and .xlsx exports. What NotebookLM can't do: - It edits your files and saves the result as a new copy. Your original is never touched. Word edits come back as tracked changes and comments. It updates Excel values and formulas, and replaces text on slides. - PDF tools: fill and flatten forms, add watermarks and page numbers, and merge, split, rotate or reorder pages. - It analyzes CSV and Excel files and charts the results. - It reads scanned PDFs and images with OCR (Latin, Chinese and Japanese script). It also transcribes audio recordings on your machine, so you can search and cite them. - MCP plugins: the agent can use MCP servers from our plugin registry, such as Notion, or any server you add by URL. They stay off until you allow them, and you choose which tools it can use and which ones ask first. - It works with any model. Coming in 2.2.3: you will be able to sign in with a Claude Pro or Max plan through your own Claude Code, in both Q&A and Agent chats. Notion is joining DeepWiki in the plugin registry. Also scanned PDFs and photos will also be read faster, since OCR goes first and Docling only runs when its needed. Where it's worse: - No video overviews. - You can't paste a YouTube link or a web page as a source. Sources are files and folders: PDF, Office files, CSV, HTML, Markdown, text, images and audio (MP3, WAV, FLAC, OGG). - Plugins only work in Agent chats, not in Q&A. - Turning Word files or slides into PDFs needs LibreOffice. It's a separate download that you turn on in Settings. Your files stay on your disk. There's no telemetry, and the only limit on how many sources you add is your disk space. MODSetter/SurfSense

Top 5 projetos open source do GitHub | Vol. 6 #shorts Os projetos #5 a #1 do Top 10 open source do GitHub, Vol. 6, 10 segundos para cada um. #5 Haven: ancsemi/Haven #4 Baserow: baserow/baserow #3 YuE2: multimodal-art-projection/YuE #2 SurfSense: MODSetter/SurfSense #1 OpenPost: getopenpost/openpost Vídeo completo com os 10 no canal Hallen Tech. Oferecimento: Hallen Tech, hallentech.com #shorts #opensource #github

10 projetos open source do GitHub que valem o seu tempo | Vol. 6 10 projetos open source do GitHub que valem o seu tempo, em contagem regressiva e com 10 segundos para cada um. Tem alternativa ao Discord, ao NotebookLM e ao Airtable rodando na sua máquina, música completa gerada por inteligência artificial aberta, VS Code no terminal e mais. 📺 Sexto volume da série Open source do GitHub. Todos com commit nas últimas duas semanas (conferido em 13/10/2026). Segue o canal e ativa as notificações para não perder a próxima. Qual desses você vai testar primeiro? Conta nos comentários! ⏱️ CAPÍTULOS E LINKS 0:00 Abertura 0:10 #10 Terminal Code: o VS Code inteiro dentro do terminal zenbu-labs/terminal-code 0:20 #9 CyberStrike: agente de IA para testes de invasão (use só em sistemas que você tem autorização para testar) CyberStrikeus/CyberStrike 0:30 #8 Android Reverse Engineering: skill do Claude Code que descompila apps Android e mapeia as APIs SimoneAvogadro/android-reverse-engineering-skill 0:40 #7 OpenViking: banco de contexto para agentes, com memória, conhecimento e skills em pastas volcengine/OpenViking 0:50 #6 TestSprite: testa o seu app como usuário real e mostra ao agente o que quebrou (a CLI usa uma chave da plataforma TestSprite) TestSprite/testsprite-cli 1:00 #5 Haven: alternativa privada ao Discord no seu próprio servidor ancsemi/Haven 1:10 #4 Baserow: alternativa open source ao Airtable, sem código baserow/baserow 1:20 #3 YuE2: músicas completas, com voz e arranjo, a partir da letra multimodal-art-projection/YuE 1:30 #2 SurfSense: alternativa ao NotebookLM que roda no seu computador MODSetter/SurfSense 1:40 Oferecimento Hallen Tech 1:54 #1 OpenPost: alternativa open source ao Buffer, Canva e CapCut getopenpost/openpost 2:04 Encerramento 🧡 Oferecimento: Hallen Tech, liderança de tecnologia sob demanda, software, cloud e segurança para a sua empresa. Site: hallentech.com E-mail: contato@hallentech.com Instagram: instagram.com LinkedIn da Hallen Tech: linkedin.com LinkedIn do Eddwan: linkedin.com Se curtiu, deixa o like, segue o canal e ativa as notificações para a próxima lista. #opensource #github #inteligenciaartificial

准备把公司资料交给本地 AI？先看文件怎么生成。 SurfSense 当前文档写明：发起 DOCX/PPTX/XLSX/PDF 生成后，会运行模型写的 Python，没有沙箱，也不另问代码执行许可；这些代码能自行联网。 先用隔离环境、公开资料试跑，再考虑内部资料。 github.com/MODSetter/SurfSense/blob/main/docs/arc… github.com/MODSetter/SurfSense/blob/main/docs/arc… t.co