Source量子位•Stalecollected in 82m
Avernet Opens Multi-Agent Collaboration to All
#multi-agent#agent-orchestrationavernetant groupavernet
Explore an open-source foundation for coordinating humans and multiple AI agents.
30-Second TL;DR
What Changed
Ant Group has officially open-sourced Avernet.
Why It Matters
An open-source collaboration layer could lower the barrier to building systems where multiple agents coordinate on complex tasks. It may also encourage more standardized patterns for human-agent and agent-agent workflows.
What To Do Next
Clone the Avernet community edition and prototype a workflow that assigns subtasks to multiple agents with human approval checkpoints.
Who should care:Developers & AI Engineers
Key Points
- •Ant Group has officially open-sourced Avernet.
- •Avernet focuses on infrastructure for multi-agent collaboration.
- •The community edition is now available for developers to explore.
Deep Insight
AI-generated analysis for this event — not the original article.
Enhanced Key Takeaways
- •Avernet utilizes a hierarchical architecture designed to manage complex task decomposition, allowing agents to operate with specialized roles rather than as monolithic entities.
- •The platform integrates a proprietary 'Agent-to-Agent' (A2A) communication protocol that optimizes latency and context-sharing efficiency compared to standard RESTful API approaches.
- •Ant Group developed Avernet to address the 'coordination tax' in large-scale agent systems, specifically targeting the challenges of state synchronization and conflict resolution in multi-agent environments.
- •The community edition includes a visual orchestration dashboard that allows developers to monitor agent workflows, decision-making logs, and resource allocation in real-time.
- •Avernet is built to be model-agnostic, supporting integration with both proprietary LLMs and open-source models through a standardized adapter layer.
Competitor Analysis
Architecture
- Avernet
- Hierarchical/Organizational
- Microsoft AutoGen
- Conversational/Peer-to-Peer
- LangGraph
- Graph-based/Stateful
Primary Focus
- Avernet
- Enterprise Coordination
- Microsoft AutoGen
- Rapid Prototyping
- LangGraph
- Complex Workflow Control
Pricing
- Avernet
- Open Source (Apache 2.0)
- Microsoft AutoGen
- Open Source (MIT)
- LangGraph
- Open Source (MIT)
Benchmarks
- Avernet
- High-scale task throughput
- Microsoft AutoGen
- High flexibility
- LangGraph
- High state reliability
| Feature | Avernet | Microsoft AutoGen | LangGraph |
|---|---|---|---|
| Architecture | Hierarchical/Organizational | Conversational/Peer-to-Peer | Graph-based/Stateful |
| Primary Focus | Enterprise Coordination | Rapid Prototyping | Complex Workflow Control |
| Pricing | Open Source (Apache 2.0) | Open Source (MIT) | Open Source (MIT) |
| Benchmarks | High-scale task throughput | High flexibility | High state reliability |
Technical Deep Dive
- Core Architecture: Implements a multi-layered framework consisting of an Agent Orchestration Layer, a Communication Bus, and a Persistence Layer for state management.
- Communication Protocol: Uses a custom message-passing interface that supports asynchronous event-driven interactions, reducing overhead in high-concurrency scenarios.
- State Management: Employs a distributed state store that ensures consistency across agents during long-running, multi-step task execution.
- Integration Layer: Provides a plugin-based system for connecting external tools, databases, and APIs, utilizing a standardized schema for tool definition and execution.
Future ImplicationsAI analysis grounded in cited sources
Avernet will become the standard for internal agent orchestration within the financial services sector.
Ant Group's focus on high-security, high-reliability infrastructure aligns with the strict compliance and operational requirements of financial institutions.
The platform will see rapid adoption in supply chain management automation.
The hierarchical coordination model is uniquely suited to solve multi-party logistics problems where agents represent different stakeholders with conflicting objectives.
Timeline
2025-03
Ant Group initiates internal development of Avernet to streamline cross-departmental agent workflows.
2025-11
Avernet reaches internal production maturity, managing over 10,000 concurrent agent interactions.
2026-08
Ant Group officially releases the Avernet community edition to the public.
- 2025-03Ant Group initiates internal development of Avernet to streamline cross-departmental agent workflows.
- 2025-11Avernet reaches internal production maturity, managing over 10,000 concurrent agent interactions.
- 2026-08Ant Group officially releases the Avernet community edition to the public.
Weekly AI Recap
Read this week's curated digest of top AI events →
Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: 量子位 ↗
This is a summary, not the original. Read the source, or get the weekly briefing.
The weekly digest
One email a week. Unsubscribe anytime.