March 12th, 2026

In the past month our updates were a bit quiet as we’ve been heads down completely rebuilding our database and search infrastructure for a consistent search experience across all current and future touch points.
This update was massive migration from Firebase to Supabase touching every part of our codebase, so please reach out asap if you are running into any issues whatsoever.
PS: If you are an existing subscriber and have not migrated to Memex v2, please reach out so i can port your subscription.
~ Oliver
You can search your saved Memex content from anywhere with full text and semantic search
Mobile App (already in review)
Obsidian Integration (in development)
N8N plugin (coming soon)
You can now use the API, MCP, Claude/OpenClaw skills to save urls into your knowledge base, see documentation linked in previous point.
Supports public websites, tweets, YouTube videos, images and you can also add tags to everything you save via the API.
This is just the beginning of making Memex’ search really good and we have many improvements planned (e.g. entity recognition, graph retrieval, better chunking etc). Please reach out with concrete examples of searches that don’t fullfil the way you want it so we understand the limitations and can improve on them
Cool side effect of this work: Moving to Supabase will open the path to self-hostability on an open source infrastructure. Not promising anything at this point there are other priorities but its possible.
Now we have dedicated support not just as an MCP/API endpoint but for Claude Plugins, OpenClaw skill and al other agent frameworks that can read skills and mcp endpoints.
Visit our docs to get started »
Much faster Youtube transcript fetching and summarisation
Showing date of saved items in results and adjust time formats via the settings

Unifying profile picture and social source flags

Speed improvements to search
Some design improvements for cleaner design
