CARV, a blockchain data protocol designed for gaming and AI projects, announced Thursday that it is launching a $50 million accelerator program aimed at fostering innovation in decentralized data infrastructure. The initiative, called CARV Labs, will support startups working on projects that could drive mass adoption of CARV's data protocol.

Backed by prominent blockchain venture capital firms including HashKey Capital and Ethereum software firm Consensys, CARV Labs will offer funding, technical support, and industry connections to selected projects. (Disclosure: Consensys is one of 22 investors in an editorially independent Decrypt.)

"Our goal is to enable a decentralized data ecosystem for broader innovation," said CARV co-founder Victor Yu, in a statement. "Building sustainable data infrastructure hinges on creating a dynamic data flywheel, which requires modular infrastructure and a critical mass of high-quality applications post-product market fit."

CARV Labs' first incubated project, an idle game called Banana built as a Telegram mini app in The Open Network (TON) ecosystem, has attracted 10 million players since launching in late July. Banana lets players tap to earn points, as well as unlock a variety of colorful and unique fruit collectibles—akin to a popular but unrelated PC game of the same name.

The accelerator has a number of gaming-centric partners, CARV said, including Netmarble subsidiary Marblex, Intella X, Doublejump.tokyo, and Xterio.

CARV says that it has generated over $5 million in revenue year-to-date from applications using its protocol. In August, CARV launched its Alphanet, operated by 40,000 community-run verifier nodes valued at $35 million.

The accelerator announcement follows CARV's $10 million Series A funding round in April, led by Tribe Capital and IOSG Ventures. The company, which has raised $50 million to date, integrates with over 900 games and AI companies, serving 9.5 million registered players across more than 40 blockchain networks.

Editor’s note: This article was written with the assistance of AI. Edited and fact-checked by Andrew Hayward.