what's Reppo?
Reppo is a decentralized protocol that provides on-demand AI training data by leveraging prediction market mechanisms to crowdsource high-quality human feedback. Built to address DePIN challenges like verification, it enables anyone to spin up subnets with minimal effort—requiring just a few clicks and front-loading REPPO tokens—to collect targeted data signals for AI models. The network optimizes data aggregation through incentives, AI agents for processing, and tokenization for financialization. Originating from insights gained at Protocol Labs on Filecoin, Reppo uses mechanism design and cryptography to solve scalable data provision for AI ecosystems, powering agents on platforms like Virtuals Protocol while creating economic opportunities for participants.
Reppo features bribing, disagreement, and shorting for voters on raw data.
Added to ClawIndex as one of 40 new apps included in the February 16, 2026 update, positioned as a network for AI training data.
Transitions to agent-native environment with @reppodant handling autonomous publishing, content routing into subnets, and persistent operations for reduced manual coordination.
Identified as an early expression of a security or data infrastructure trade alongside zauth and oso, according to a CT commentator's investment thesis.
Shipped three new subnets and expanded beyond AI training data.
Publishers and voters earned over 75,000 REPPO tokens.
Introduces bribing mechanics and disagreement/shorting features for voters on raw data.
Top publishers can earn up to approximately $2,000 every two days if their data performs well within the platform's prediction market and subnet structure.
Brings back robotics subnets on 2026-02-01 using wearable cameras and prediction markets for crowdsourced robot training data.
formed partnership with Starlythestar for city-scale activations
became the first customer of Starlythestar's "Proof of Presence" feature which is now live
Uses its token to incentivize data labeling through a staking mechanism where users lock tokens, review and label data, and receive rewards only for quality judgment.
Achieved over 100 million votes processed on the network within the first month after launch.
Has 13 live subnets operational on the network.
Has paid out $14,000 to users since launch.
Holder count is steadily climbing.
Maintained 67% relocking rate.
Unveils remixing feature for on-chain IP with provenance and monetization, leveraging ERC-7053 for raw data reuse across subnets, with rollout scheduled within 4 weeks.
oracle_opz seeded a subnet on reppo.
Market cap reaches approximately 34M, up from the 19M FDV reported on December 4th.
@epochbiz schedules research article publication about the project for today.
REPPO ranks #5 among top traded tokens on Wasabi this week (excluding majors).
Token reaches new all-time high following apology from Virtuals core contributor.
Virtuals protocol founders publicly tweet about and endorse the project, with current FDV at 19M.
Wasabi Earn Vault attracts 1.23% of token supply with 57% APY while price increases 247%.
REPPO token lists on Wasabi platform with 3x leverage trading and yield earning opportunities.
Project raises $2.2M at $30M valuation and launches token at $2M market cap with seed investors locked for 6 months. EigenLayer founder follows the project.
New airdrop campaign announced for the project, which is positioned as an AI training data network.