@Anchorage@paradigm@TakoViz MPP is an elegant stablecoin-based solution to a real problem for humans, agents, and companies like us building for them.
Great work from @gakonst , @jeff_weinstein , and everyone else pushing this forward!
MPP is a full-circle moment for me. I joined @Anchorage and @paradigm early because I was fascinated by reinventing the infrastructure for how value moves.
Tako now supports the Machine Payments Protocol (MPP), so your agents can now pay for real research.
It's simple: just set your agent up to pay with Stripe or stablecoins via Tempo and now your agent can autonomously discover and pay for Tako’s APIs, unlocking real-time, authoritative answers across markets, startups, web + app analytics, public opinion, geopolitics, sports, weather, and even your own data.
MPP removes even more of the friction separating your questions from great answers. Check it out!
I eventually left crypto to work on another infrastructure problem: giving AI access to high-quality answers grounded in real-time data.
And in working on that at @TakoViz, I ran into a related one: how do agents pay for answers without a human in the loop? And how do you make paying fractions of a cent actually practical?
Excited to partner with @TakoViz!
Tako turns natural language questions into interactive visual analytics. Ask "Nvidia vs AMD revenue since 2015" and get back a polished, cited chart from S&P Global. Covers stock prices, crypto, prediction markets, polling, weather, sports, and more.
An AI data analyst your agents can call with one API request.
Now on Orthogonal via Skills, MCP, SDK, or CLI.
Adish, you nailed it! Consistency and correctness are hard. Rather than generate from scratch we have an ontology that codifies our understanding of data + a library of components. The model assembles components into "cards" and uses ontology to ensure the card and its contents make sense. Its very different from text in -> text out!
We had a lot of fun building the Tako-powered Research Canvas App! CopilotKit and LangChain made assembling an agentic workflow really easy. The langgraph tooling is great. Agents consuming/assembling UI components via MCP is the future 🚀
🦜🐙 @LangChain + @tavilyai + @TakoViz and CopilotKit
Generative UI Research Canvas
A research assistant that dynamically creates beautiful, interactive reports.
How it works:
- LangChain for agent complex workflow
- Tako to generate dynamic visuals from public structured data sources
- Tavily search assistant
- CopilotKit for the frontend that enables user-interactivity
This is more than just a research canvas-
It's a blueprint for how agents + Generative UI can create a powerful UX
It's open-source!
📗 Tutorial: https://t.co/osVOGjy58C
👾 Repo: https://t.co/HyW1xVLVHM
Docs: https://t.co/6ZQZXjAKCX
@TakoViz@CopilotKit We had a lot of fun building the Tako-powered Research Canvas App! @CopilotKit and @LangChain made assembling an agentic workflow really easy. The langgraph tooling is great. Agents consuming/assembling UI components via MCP is the future 🚀
Le Chat just got smarter. 🐙
Tako now powers real-time data and visualizations inside @MistralAI Le Chat.
→ Accurate answers
→ Live, citable data
→ Engaging visualizations
Check it out!
1/ LLMs made intelligence a metered utility; you now get reasoning on demand. Metered intelligence requires everything upstream to meter itself too: data, context, etc.
That's why web search APIs are suddenly becoming the next major component of the “intelligence grid."
🐙 We’ve added another world-class data provider to Tako’s knowledge graph: @Similarweb
Now your agent can ask web and app performance questions like:
• “How much traffic does cursor receive in France?”
• “chatgpt bounce rate”
And receive real answers - not vibes.
Check it out, and let us know what you build!
And, the "up and coming" products I'm currently testing and excited about:
- Serif (@serif_ai) - email assistant based on your knowledge base
- Tako (@takoviz) - AI engine for trusted data with beautiful graphics
- Grok (@grok) - image/video gen models on mobile are amazing
- M1 - AI receptionist. Most users are SMBs but I'm fascinated by the concept for consumers
- Genspark (@genspark_ai) - horizontal agent, I need to try out more workflows on this!
- SnapCalorie (@snapcalorie) - AI nutrition tracker on mobile
- Rememe (@rememe_co) - AI social photo generator (via face swap)
We're thrilled to partner with @GnomiApp - a News AI Agent that delivers accurate, real-time news tailored to you.
Tako's real-time knowledge search + visualization capabilities are now live for Gnomi's users. Our information-dense, visually engaging knowledge cards surface contextually relevant insights that help users grasp stories in seconds.
We're excited to build the future of news together. Check it out!
We carefully designed the Tako Knowledge Search API to be very simple to understand and use which makes it easy for developers, agents, and vibe coders.
🤯 Really impressed by this slick, vibe-coded (!) app 📈 📊 that one of our users created.
👉 https://t.co/KKh824FGHh