logo
DocsBlogCareersAPI
discordxtelegram
Multi-Agent
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
A
A
B
A
Layer

Improving Multi-Agent Coordination in Decentralized Ecosystem
Backed by Google for Startups Cloud Programgoogle

Join Community
01
engage
The orchestration LLM connects agents and users to act smarter, together.
02
rank
Compare, rate, and reward - Users evaluate agent responses to drive self-optimizing LLMs toward a smarter outcome.
03
build
A composable framework helps to build your agent in your preference, with Bonus.
block
01 engage
Engage Smarter with Supernet
The Orchestration Framework provides a unified API to seamlessly connect agents and users. Supernet eliminates fragmented tools in a single platform and deploys a multi-agent optimizer (yes, it’s an agent too!) to activate multi-agent collaboration to better fit users intent.
02 rank
Rate, Reward and Optimize in Supernet
Users do an easy evaluation of agent responses. Boosted by Web3 incentive models on feedback, user contributions are incentivized, accelerating LLM self-iteration through adaptive learning loops, automated hyperparameter tuning, and real-time performance analytics. This creates a dynamic, self-improving performance model that delivers faster, more accurate results with every iteration.
03 build
Easy Build with Supernet
Build agents with our flexible framework that seamlessly integrates your own data/GPU/API with our modules. Supernet handles dirty work like RAG setup, iteration, and tuning to ensure builders focus on innovation. Plus, Supernet provides bonuses for builders who contribute to and optimize our growth.
Web3 and AI converge to
create decentralized
1payment
Payment
  • No currency barrier and restriction
  • Protect payment information
2payment
Incentive
  • Incentivize high-quality agents to gain more exposure.
  • Encourage users to participate in the growth of the DAO ecosystem
3payment
Transparent
  • Transparent ranking and genuine market feedback.
  • Open source Multi-Path Recommendation Algorithm
section6
0
AI iteration count
section6
0
Total agent invocations
section6
0
Points consumed
arrows
arrows
Build with Us
Join Community