I've spent some time studying the current version of the AI Agent framework Eliza from AI16Z and the differences between the current version and the V2 version that Shawn has revealed. Before comparing them, let's first understand the basic architecture of the current version and the V2 version.
As shown in Figure 1, the current version of the Eliza framework consists of the Interface Layer, Core Layer, and Model Layer.
Interface Layer
- Responsible for interacting with external platforms
- Supports multiple platforms such as Discord, Twitter, and Telegram
- Provides an API interface for other systems to call
Core Layer
- Memory system: Provides searchable storage and file processing capabilities
- Behavior system: Manages custom behaviors and action execution
- Configuration system: Manages configurations through .env and elizaConfig.yaml
Model Layer
- Supports various large language models, including OpenAI, Anthropic, Llama, Grok, and others
- Provides a unified model interface abstraction
The layers interact with each other through standard interfaces, supporting a high degree of modularity and extensibility.
On the 15th, Shawn revealed that they are developing the Eliza V2 version, and an alpha version is expected to be released in 2-3 weeks.
Eliza V2 will unify the message bus and simplify the client; unify the Agent wallet; adopt a registry and override model system; enhance the extensible and generic core framework; update the community plugins; and achieve 100% test coverage.
The estimated architecture design of the V2 version is as follows:
Let me explain the main improvements in Eliza V2 and their positions in the architecture:
Improvements to the Interface Layer: Unified Message Bus
- Integrates the previously scattered clients into a unified message processing system
- Each platform (Discord/Twitter, etc.) becomes an interface card for the message bus
- Simplifies client development and maintenance
- Provides a more consistent message processing experience
Improvements to the Core Layer: Unified Agent Wallet
- Integrates the previously scattered wallet configurations
- Provides a unified multi-chain wallet management
- Simplifies cross-chain operations
Extensible Core Framework: Redesigning the Core Interface to be More Versatile
- Enhances the extensibility of the framework
- Improves the plugin integration mechanism
Community Plugin Updates: Improving the Plugin System
- Enhances plugin management
- Ensures 100% test coverage to guarantee quality
Improvements to the Model Layer: Introducing a Model Registry Mechanism
- Supports dynamic model registration and overriding
- Categorizes models into two main types: LLMs and Embeddings
- Provides more flexible model invocation methods
This new version of the architecture is more modular and unified, with clearer interactions between different components, and provides a better foundation for future expansion. Finally, from an investment perspective, Eliza V2 will bring new growth opportunities for plugin extensions like AgentTank.
