Gonka's total network computing power exceeds 10,000 H100 equivalents, with the five major inference models used nearly 100 million Tokens in a single day

PANews December 17 News, according to Gonka browser data, the decentralized AI inference network Gonka’s total network computing power has surged nearly 20 times, surpassing 10,000 NVIDIA H100 equivalents, reaching 10,729 (as of December 17). This scale is comparable to a large national-level AI computing center or a leading cloud vendor’s core AI cluster, capable of supporting continuous high-throughput inference for models with hundreds of billions of parameters and training for models with trillions of parameters simultaneously. Gonka integrates global GPU resources through a decentralized approach, forming a sustainable, high-performance AI computing network without the need for centralized data centers. This computing milestone marks Gonka’s official entry into the ranks of top-tier global AI infrastructure networks, with the capability to provide commercial-grade API services. Currently, Gonka supports five major AI inference models. In just three months since its mainnet launch, the total daily (per epoch) inference model usage has approached 100 million tokens. The mainstream model Qwen3-235B-Instruct is used approximately 30 million tokens daily, showing exponential growth potential. There are nearly 600 active nodes from over 30 countries and regions joining the Gonka computing network, with more than 2,000 users utilizing AI inference services daily. Data indicates that Gonka’s inference usage growth has far outpaced the growth of network nodes and computing power, validating the market’s rigid demand for Gonka’s decentralized inference services and the viability of its business model. According to previous reports, as an efficient decentralized AI inference and training network, Gonka recently received a $50 million investment from Bitfury and is supported by investors such as Coatue, an investor in OpenAI, and Slow Ventures, an investor in Solana. It is regarded as one of the most promising emerging infrastructures in the AI × DePIN direction.