Real customers. Real revenue. Coming fully on-chain this month.
ReadyAI Update and Roadmap:
→ 5 paying customers including a NYSE-listed REIT (SmartStop, $1.8B)
→ x402 payment rails going live May 18th with every API call settable and verifiable on-chain
→ 75% of enrichment revenue committed to SN33 alpha buybacks
→ Public dashboard shipping this month so anyone can audit demand
Full breakdown ↓
When we started @ReadyAI_ the thesis was simple: structured data is the bottleneck for AI, and Bittensor is the best way to produce it at scale. Today we can prove it. NYSE-listed customers paying for our data, those payments moving on-chain through x402, and 75% flowing into alpha buybacks. Dashboard goes live next week.
We've been heads down on something. Coding agents are the trillion-dollar race for every major lab, and the high-quality structured data to make them work is the bottleneck. Context7 became the #1 MCP server (53K stars) solving current docs. But the hard problems live deeper from version-pinned breaking changes to expert reasoning mined from thousands of technical podcasts, coding intelligence that doesn't exist in any documentation. We're building that dataset. More this week.
If Martin is right, he also just wrote the product spec for open source + distributed compute where broad swaths of groups, individuals and organizations contribute their compute resources to training runs for large param open source models.
There are lots of issues in figuring this out: homogeneity vs heterogeneity of the training clusters, orchestration, financial incentives etc etc etc but some early projects are good signal as to where this can go and that these limitations can be overcome (folding@home, Venice, Tao).
An attempted oligopoly on intelligence is the perfect boundary condition for a bottoms up uprising of fully open, fully distributed AI.
New on @ReadyAI_
Request an llms.txt file for any domain, free
Search a site → not in our 10K+ database? Hit "Request This Domain Now" → get your file queued on subnet
5 free requests per user. Every file is open-sourced on GitHub
Structured data for agents shouldn't be gated
This is the best explanation for why
(1) Bittensor is unique amongst crypto projects and
(2) you often see crypto VCs hating on it
Bittensor provides the incentives for bootstrapping innovation across numerous experiments all at once without the need for VCs
$TAO
Yes emissions are used to bootstrap innovation, same as Uber, Amazon and countless of other big companies
You can chose between these 2:
-give those emissions to VC’s
-give those emissions to builders who devote their whole time to build out the network
Vc’s hate it because they can’t apply the VC playbook/had discounted access compared to the masses
We just launched a new https://t.co/MoC5WHinEv
Type any domain into the search. If it's in our dataset, you get clean, structured intelligence instantly. No scraping. No parsing HTML. Just machine-readable data, ready for any AI agent.
10,000+ websites crawled, cleaned, and structured by Subnet 33 so far. Growing to 100K by Q2, 1M by year end.
This is the beginning of something bigger: a marketplace for agentic data.
Right now, every AI agent that needs info about a company or domain scrapes, parses, and hopes. Billions of redundant crawls. Trillions of wasted tokens.
We're building the infrastructure layer that fixes this — an indexed, machine-readable web powered by decentralized compute.
Our recent breakthrough with enrichment tasks on the subnet has completely opened the floodgates. We can now create structured datasets from nearly any source, from llms.txt to deep coding data.
Will be sharing benchmark improvements with this coding data shortly
If the number of agents grows explosively and each one has to crawl the web independently, the resulting compute consumption will be enormous.
The hidden opportunity here is to convert “webpages → AI-readable structured data,” which avoids massive redundant compute usage.
That’s why I’m bullish on what SN33 @ReadyAI_ & @DavFields is building — The Data Layer
The web wasn't built for AI agents. We're fixing that. First 1,000 domains live now, millions coming. Open source, decentralized, and free.
Frontend coming shortly to request llms.txt for any site
🚀 llms.txt are live on SN33
The llms.txt repository is now live. 🔗 https://t.co/fN2K02fBLl
SN33 has processed the first batch with over 1,000 websites crawled, cleaned, and converted into structured llms.txt files by the subnet.
Semantic summaries ready for any LLM agent, MCP server, or AI app to consume instantly. No scraping. No parsing raw HTML. Just clean, machine-readable intelligence.
New batches will be pushed as the subnet keeps processing. The repo grows every week.
What's in the dataset:
→ Structured semantic summaries per domain
→ Named entities: people, orgs, products, technologies, concepts
→ Topic classification and key themes
→ Deterministic O(1) lookup by domain with no index file needed
→ Git-friendly structure that scales to millions of domains
This initial release covers ~1,000 domains as a pilot, but the pipeline scales to millions.
📍 Roadmap: 10K → 100K → 1M domains → continuous updates from new Common Crawl releases and soon from requests.
🌍 And the frontend is coming.
Any domain. You request it, the subnet processes it, you get an llms.txt back. We're putting the finishing touches on the public UI and it drops soon.
SN33 is becoming infrastructure. The web, made readable for machines and open to anyone, powered by decentralized infra.
Star the repo. Share it. And stay close. The next drop is right around the corner.
We just completed the largest decentralised LLM pre-training run in history: Covenant-72B. Permissionless, on Bittensor subnet 3.
72B parameters. ~1.1T tokens. Commodity internet. No centralized cluster. No whitelist. Anyone with GPUs could join or leave freely.
1/n
1000+ websites processed. Open-source repo Thursday. Public frontend right after. SN33 is becoming the infrastructure layer between the open web and the agent economy. Onward!
👀 something new is coming
We've been building and we're almost ready to show you.
SN33 has been processing the web at scale, turning raw Common Crawl data into clean, AI-ready `llms.txt` files. Structured semantic summaries that any LLM agent, MCP server, or AI app can consume instantly.
On Thursday we'll be releasing the Github repo where `llms.txt` files will be pushed in batches as the subnet processes them. We're starting with over 1000 websites analyzed and processed by the subnet that will grow every week.
And shortly after...
🌍 We're launching a public frontend
Any website. Any domain. You request it, the subnet processes it and you get a `llms.txt` back.
No more raw HTML hell for AI agents. No more redundant crawling. Just clean, structured, machine-readable intelligence about any corner of the web, on demand, powered by decentralized compute.
This is SN33 becoming a public utility for AI infrastructure
The web, made readable for machines. At scale. Open to anyone.
🔜 More very soon. Stay tuned.
@DallasAptGP If they are doing $200 mil ARR with 100k users that is $2k per user PER year or $166 per user per month. Much closer to Claude premium plan. Do you know which it is?
The generic data race is over. The teams that win the next 3 years are the ones building deep, vertical-specific pipelines that scraping can't replicate.
That's exactly what we're doing @ReadyAI_ . Phase 1 is just the start.