Perceptron Network and BlockMesh announced in June 2025 that BlockMesh had merged into Perceptron, presenting the combination as a way to collect and refine data for artificial intelligence using consumer devices and community contributors. A formal press release went out from Dubai on June 20, 2025, a day after the news first appeared on BlockMesh's own newsletter.
Historical news event: 2025-06-20. Retrospective report prepared November 2026.
"Today, we're excited to announce that BlockMesh has officially merged into Perceptron Network," the June 19 post began. It described the deal as "a full-scale integration" rather than a name change. The June 20 release, distributed by Chainwire, described the result as the first end-to-end decentralized AI data infrastructure. The BlockMesh newsletter has since been taken offline, so the original post survives through the Internet Archive.
Why the two projects fit together
Perceptron's pitch starts with a problem AI developers know well. Training and running large models takes huge amounts of useful data, and gathering material that is current, relevant and legally usable is expensive. "We're unlocking the world's most valuable resource -data- by putting the power back in the hands of everyday people," co-founder Peter Anthony said in the June 20 release.
BlockMesh brought the collection layer: a network of devices that crawl public websites, APIs and documents and structure what they find. Perceptron brought what it called an "AI-native" framework, with agents that interact with users, classify content and label data at the edge. The merger post described this as "a dual-engine decentralized system" of web nodes and AI agents, with the agents also tagging conversational data from platforms such as Discord and Telegram.
In the companies' telling, edge agents capture context, web nodes gather and structure public data, and clean, labeled data then flows directly into AI systems. It's a familiar idea from decentralized physical infrastructure (DePIN) projects, applied here to AI data rather than storage or wireless coverage. Our earlier pieces on how blockchain can help artificial intelligence and crypto projects involving AI trace how long this pairing has been discussed.
The numbers, and who supplied them
The announcement leaned on scale. The companies cited more than 700,000 devices connected to the BlockMesh network and more than 200,000 users reached through Perceptron's AI agents. They also claimed operating costs "90% cheaper than centralized alternatives" and said they had spent nothing on traditional advertising or paid influencers.
All of those figures came from the companies. They aren't independently audited counts of continuously active nodes or unique people, and the cost comparison would need a like-for-like benchmark before it could be taken at face value. The U.Today page that carried the release labels it sponsored content and says the outlet doesn't endorse its claims.
The project also highlighted PERCs, which it described as "proof-based, trust-layer NFTs" that reward human input. An NFT badge can record that a contribution happened. On its own it doesn't prove a data sample is accurate, or that a contributor is owed a particular payment.
What it meant for BlockMesh node operators
For people already running BlockMesh nodes, the practical message was continuity. The merger post said a full snapshot of BlockMesh activity had been taken. "All contributions will be recognized and rewarded," it said. Node operators were told they would earn based on real engagement and validated AI training work, with early access to rewards, priority staking and dashboard tools. It promised "community-first reward cycles" but didn't publish a token schedule.
That's a common pattern in points-based crypto networks: contributions are logged now, with the form and value of any later token left open. Readers should treat such promises as plans, not guaranteed payouts.
Quality and consent
A distributed network can collect information from many locations, which helps with websites that show different content in different countries. Geographic reach brings its own problems, though: duplicated material, altered pages, privacy expectations and licensing limits. A page being publicly visible doesn't make every use of its text or images unrestricted.
The companies argued that automated validation and contributor incentives would cut down on low-quality data. Whether the combined system can deliver reliable, legally sourced datasets at scale depends on the quality of what it collects, a real paying market, transparent rules for contributors and repeat customers. The June announcement set out the intended design and ambitions. It didn't prove future revenue or network claims.
What came next
A year later, Perceptron was describing itself mainly as a data supplier. In a July 2026 interview, Anthony said the network had grown to roughly 800,000 nodes across more than 150 countries, figures again supplied by the company, as we report in Perceptron's account of turning idle bandwidth into training data. Later that month the company announced $6.5 million in strategic funding.
The bigger idea in the merger is that ordinary connected devices and people could help produce useful AI infrastructure, with a blockchain recording their contributions rather than serving as the only reason to join. Whether that works is still being tested. Others in the sector, such as 0G with its Aristotle mainnet, are chasing related goals from the compute and storage side.
Frequently asked questions
When did BlockMesh merge with Perceptron Network?
BlockMesh announced on June 19, 2025 that it had merged into Perceptron Network. A formal press release followed on June 20, 2025.
What is Perceptron Network?
Perceptron Network is a decentralized AI data project that collects public web data through a network of consumer devices and uses AI agents to label it. Contributors are rewarded through points and PERC NFTs.
How many nodes did BlockMesh have?
The companies said BlockMesh had more than 700,000 connected devices at the time of the merger. That figure was supplied by the companies and wasn't independently audited.
What happened to BlockMesh users' rewards?
The merger post said a full snapshot of BlockMesh activity had been taken and that all contributions would be recognized and rewarded under Perceptron's system. It didn't give a token schedule.
Sources
- BlockMesh newsletter, June 19, 2025 merger post (archived)
- Chainwire press release via U.Today, June 20, 2025
- Perceptron Network on X
- Crypto.news, July 3, 2026 interview with Perceptron CEO Peter Anthony
Featured image: Perceptron Network
Frequently asked questions
When did BlockMesh merge with Perceptron Network?
BlockMesh announced on June 19, 2025 that it had merged into Perceptron Network. A formal press release followed on June 20, 2025.
What is Perceptron Network?
Perceptron Network is a decentralized AI data project that collects public web data through a network of consumer devices and uses AI agents to label it. Contributors are rewarded through points and PERC NFTs.
How many nodes did BlockMesh have?
The companies said BlockMesh had more than 700,000 connected devices at the time of the merger. That figure was supplied by the companies and wasn't independently audited.
What happened to BlockMesh users' rewards?
The merger post said a full snapshot of BlockMesh activity had been taken and that all contributions would be recognized and rewarded under Perceptron's system. It didn't give a token schedule.
Last updated NOV 12, 2026
