Search Everything.
Leak Nothing.
Pay Once.
Press Ctrl+Alt+Space to search Obsidian, Paperless, Nextcloud, your email, and four more sources at once — powered by your own local LLM. No cloud round-trip. No subscription. Your data never leaves your machine.
No credit card or email required. Installs in seconds.
Stop Digging. Start Asking.
Information is scattered across your notes, wikis, local PDFs, and self-hosted apps. Prism bridges the gap. A single query searches all your tools at once, synthesizes the context and hands you the answer — every result labeled by the source it came from.
Vector cache — instant semantic matches from previous searches
InstantScanned documents, receipts, PDFs with full OCR text
LiveWiki pages, shelves, chapters — your documentation hub
LiveMarkdown notes, daily journals, knowledge graph links
LiveSearXNG — private web results without tracking
LiveLocal LLM chat — run open models on-device via LM Studio or any OpenAI-compatible server
LiveFull-text search across local documents, code, and text files
LiveGmail, Microsoft 365 and FastMail search — find messages without leaving the palette
LiveSelf-hosted cloud — search files, calendars, and contacts
LiveBYOK cloud AI — frontier models on your terms, your API keys
LiveQdrant cached results appear instantly. Live sources stream in as they complete.
Your Entire Digital Brain, One Keystroke Away
Not just search. Prism is a full AI command center that turns scattered knowledge into instant, cited answers — without ever leaving your machine.
Universal Search
One query, eight sources, zero context switching. Search Obsidian, Paperless, Nextcloud, local files, and more in parallel. Results stream in labeled by source, so you always know where an answer came from.
Local or Cloud, Your Call
Run models locally through LM Studio or any OpenAI-compatible server, or bring your own key for Claude and GPT. Every query routes to the model you pick — private on-device inference by default, frontier cloud power when you want it.
Private AI Chat
Have streaming conversations with your local LLM — no data ever leaves your machine. Ask follow-ups about the results you just found, with a personal AI that actually stays personal.
Take Your Answers With You
Export any summary or AI answer to Markdown, PDF or Word — grouped by source with links intact. Or copy a clean summary or the full reference list in a single click.
Smart Prompts
Type / to browse categorized prompt templates for coding, writing, analysis, and more. Skip the blank-page problem and get to work faster.
Always One Keystroke Away
Press Ctrl+Alt+Space (Windows/Linux) or Cmd+Option+Space (macOS) to summon Prism over any app. No window switching, no browser tabs. Your AI command center appears instantly, right where you’re working.
Your Knowledge, Organized
Every search is saved locally. Pin important results for quick recall. Your search history is yours alone — stored on your machine, never uploaded, always portable.
Deep Integrations
Fire webhooks on search, chat, or OCR events. Connect Prism to Zapier, Make, or any HTTP endpoint to build powerful automations that fit your existing workflow.
What Happens on Your Machine, Stays on Your Machine.
Your legal briefs, financial reports, and internal wikis are never sent to OpenAI, Google, or Microsoft. Prism runs entirely on your hardware — every search, every chat, every AI response stays on your local disk. Nothing is uploaded and there is no telemetry.
- No telemetry or usage analytics — ever
- All AI inference runs on your hardware — zero cloud calls
- No tracking pixels, no fingerprinting, no cookies
- Works fully offline — airplane mode ready
- All data stored in local SQLite — portable, inspectable, and yours
- Built on Tauri 2 + Rust — audit the code yourself
The same rule applies to this website: no analytics, no error tracking, no third-party fonts. Open your browser's network tab and check. The only requests that ever leave the app are the ones you configure — your own self-hosted services and any cloud AI key you choose to add.
Connects to the Tools You Already Use
Stop tab-switching between every app that holds a piece of the answer. Prism plugs into your existing stack — self-hosted, local, and cloud — and searches them all from a single interface.
Paperless-ngx
Document management with OCR. Search scanned receipts, invoices, contracts, and any PDF in your archive.
Self-HostedBookStack
Self-hosted wiki platform. Search across shelves, books, chapters, and pages of your team documentation.
Self-HostedObsidian
Markdown-based knowledge base. Full-text search across every note in your vault through the Local REST API plugin.
Local VaultNextcloud
Self-hosted cloud storage. Search the files on your own server through Nextcloud’s search API.
Self-HostedLocal Files
Index and search files on your local disk. PDFs, text files, code, and documents — all searchable from the palette.
LocalSearch Gmail, Microsoft 365 and FastMail from the palette — message subjects, senders and snippets via each provider’s API.
Self-Hosted / CloudSearXNG
Privacy-respecting web search. Get real web results without Google tracking. Self-host SearXNG and query it from the palette.
Self-HostedLM Studio
Run open-weight LLMs locally. Streaming chat, completions, and embeddings on your own GPU. Llama, Mistral, Phi, and more.
LocalLightweight local LLM runner. Pull and run models with a single command. Native API for chat and embeddings.
LocalAnthropic / Claude
Bring your own API key for Claude. Use frontier models for complex reasoning while keeping queries on your terms.
BYOK CloudOpenAI-Compatible
Any OpenAI-compatible endpoint — GPT, Groq, Together or a self-hosted server. Ollama’s compatible API should work too, though it hasn’t been through a verified setup pass yet.
BYOK Cloud / LocalQdrant
High-performance vector database. Semantic recall over your indexed Paperless, BookStack and Obsidian collections, embedded by your local model.
Self-Hosted