Telegram is set to become the first significant user of its own Cocoon network. This initiative aims to harness the platform’s vast user base and high query volume to establish a new economy for private AI inference on the TON blockchain.
Summary
- Telegram introduced Cocoon, a decentralized AI inference network leveraging TON.
- Developers can utilize GPU resources and pay with TON tokens; GPU owners earn by providing computational power.
- AlphaTON Capital is supporting the initiative with global GPU infrastructure for private, on-chain AI processing.
On Oct. 29, Telegram CEO Pavel Durov presented Cocoon at the Blockchain Life conference in Dubai, a decentralized network designed for the private processing of artificial intelligence queries on the TON blockchain.
Durov stated that the system would connect developers needing computational power with GPU owners willing to share their resources, rewarding them in Toncoin (TON) tokens for running inference tasks. Telegram intends to be the first major user of Cocoon, routing parts of its AI workloads through the network and promoting the platform to its extensive audience of over a billion users.
The announcement in Dubai marked the first public insight into Telegram’s AI aspirations. Durov initially hinted at the project during an October visit to Kazakhstan, where he unveiled a dedicated AI laboratory.
At that time, he disclosed that Telegram had spent months advancing technology at the intersection of blockchain and AI, which he saw as crucial for delivering confidential and transparent AI to its vast user base.
For Cocoon to operate effectively, it needs a robust participant pool. According to its official channel, the onboarding process is active. Application developers wanting to use the network can message the channel directly, specifying their intended AI model architecture, such as DeepSeek or Qwen. They must also provide details on their expected daily query volume and average token size for inputs and outputs.
On the supply side, GPU owners wishing to earn TON tokens for their computational contributions must outline their hardware capabilities. The channel requests specifics on available GPUs, including model type, VRAM size, and anticipated uptime.
AlphaTON Capital supports the Cocoon network
Following Durov’s announcement, AlphaTON Capital affirmed its strategic investment in high-performance GPU infrastructure to aid Cocoon’s implementation. The Nasdaq-listed firm, recognized for its extensive involvement in the TON ecosystem, announced it would deploy advanced GPU fleets across global data centers to support the network’s computational needs.
CEO Brittany Kaiser described the launch as a “pivotal moment in the convergence of blockchain technology, data protection, and artificial intelligence,” pledging the firm as a “cornerstone contributor” to the ecosystem. Executive Chairman Enzo Villani stressed that “privacy and security are essential in the AI era,” asserting that Cocoon’s design meets a critical market demand.
