Brain Dump

Your second brain, in your browser.

Drop in markdown and PDFs. Paste a URL. Point at a GitHub or GitLab repo. Chat with all of it, search it, and keep the keys in your pocket. Bring your own provider — OpenRouter, Ollama, Groq, OpenAI, Anthropic.

Bring your own provider

OpenRouter, Ollama (local), Groq, OpenAI, and Anthropic today — adding more is one file. You pick the model. You pay the bill.

Ingest anything readable

Files (markdown, PDF, text, code), URLs (Mozilla Readability extracts the article), and entire GitHub or GitLab repos via the API.

Runs entirely in your browser

No server. No account. Files are parsed, chunked, and embedded right here. Vectors live in IndexedDB on your machine.

Hybrid search + cited RAG chat

Vector similarity, BM25 keyword, and Reciprocal Rank Fusion when you want both. Citations are clickable — click a chip to see the source chunk in context.

Your keys and your data never leave your browser.

We don't have a server. Your API keys live in IndexedDB on your device. Your documents are parsed locally and embeddings are stored locally. Read more in our security notes.