FLock.io officially releases FLock Web3 Agent Model
ChainCatcher news, FLock.io officially releases the FLock Web3 Agent Model, which is a large model optimized for Web3 tasks, possessing precise function calling capabilities and efficiently understanding and executing complex Web3 tasks.
This model is trained by the FLock community in the AI Arena, where users can act as task creators, training nodes, validators, and delegators to jointly promote the evolution of the model. The model is based on Task1, and model contributors can receive corresponding rewards.
The FLock Web3 Agent Model achieved a 75.93% precision match rate in the web3 function calling benchmark test, surpassing GPT-4o, Gemini Flash 2.0, DeepSeek V3, and Qwen 2.5. The model and related evaluation standards are fully open-sourced, ensuring transparency and accessibility.
The FLock Web3 Agent Model has partnered with Ionet, OpenGradient, and HashKey Chain for deep collaboration to accelerate the implementation of DeAI technology and jointly expand the application boundaries of decentralized AI.