Author: Arndxt, Threading on the Edge; Translation: Jinse Finance xiaozou
There are four major frameworks in the Crypto x AI field: Eliza (AI16Z), GAME (VIRTUAL), Rig (ARC), and ZerePy (ZEREBRO).
These four frameworks can meet different development needs.
Driven by the first-mover advantage and the thriving TypeScript community, Eliza dominates with a market share of about 60%, while GAME (market share of about 20%) targets rapidly adopted games and virtual world applications.
Rig (~15% market share) is developed in Rust and provides performance-oriented modularity suitable for the Solana ecosystem, while ZerePy (~5% market share), a new Python-based framework, focuses on creative output and social media automation. The total market value of these frameworks is $1.7 billion, and as AI crypto applications expand, the total market value of these frameworks may reach more than $20 billion, making the market value weighting method potentially attractive. Each framework occupies its own unique market - social and multi-intelligence agents (Eliza), games/virtual worlds (GAME), enterprise performance (Rig), and creative community use (ZerePy) - providing complementary options to each other rather than competing directly.
（1）Eliza ($AI16Z)
● Market share: ~60%
● Market value: US$900 million
● Core language: TypeScript
● Main advantages: first-mover advantage, extensive GitHub community (6,000+ stars, 1,800 forks)
● Focus: multi-agent simulation, cross-platform social engagement
As one of the earliest AI agent frameworks in the field, Eliza has taken a dominant position. Its first-mover advantage is supported by a large community of contributors, which has accelerated development and promoted user adoption. Eliza's TypeScript stack makes it well suited for developers working in a web-based ecosystem, ensuring broad appeal.
（2）GAME (VIRTUAL)
● Market share: ~20%
● Market cap: $300 million
● Core language: (API/SDK based; language agnostic approach)
● Key advantages: Rapid adoption in the gaming industry, real-time agent capabilities.
● Focus: Procedural content generation, adaptive NPC behavior.
GAME is tailor-made for gaming and virtual world applications. Its API-driven architecture and close connection with the VIRTUAL ecosystem has generated tremendous momentum: more than 200 projects, 150,000 requests per day, and rapid weekly growth. GAME's code-free integration further attracts teams that prioritize fast deployment over deep technical customization.
（3）Rig (ARC)
● Market share: ~15%
● Market cap: $160 million
● Core language: Rust
● Main advantages: performance, modular design (enterprise level)
● Focus: "pure-play pure games" based on Solana, emphasizing retrieval-enhanced generation.
Rig is based on the Rust architecture and caters to developers who value speed, memory safety, and efficient concurrency. It is designed for "enterprise-level" or heavily data-driven applications, and is particularly suitable for applications on Solana. Despite a steep learning curve, Rig offers modular performance and reliability that could appeal to system-oriented developers.
(4)ZerePy (ZEREBRO)
● Market share: ~5%
● Market cap: $300 million
● Core language: Python
● Key strengths: Community-driven creativity, social media automation.
● Focus: Agent deployment on social platforms, particularly for artistic or niche outputs.
ZerePy is a latecomer, derived from the core backend of Zerebro. Its Python foundation, combined with a focus on creative applications (NFTs, music, and digital art), has attracted a fervent following. The collaboration with Eliza has raised ZerePy’s profile, but ZerePy’s narrow focus may limit widespread enterprise adoption.
（1）Eliza (AI16Z)
● Multi-agent systems: Deploy multiple AI personalities under a shared runtime.
● Memory management (RAG): A generation pipeline that implements retrieval enhancements for long-term context.
● Plugin system: Supports community-developed extensions for speech, text, and media parsing (e.g., PDF, images, etc.).
● Broad model support: Integrate with local open source LLMs or cloud APIs (OpenAI, Anthropic).
Eliza’s technical design is centered around multimodal communication, making it a good fit for social, marketing, or community-based AI agents. While it excels at easy integration (Discord, X, Telegram), large-scale use also requires careful orchestration of different agent personalities and memory modules.
(2)GAME (VIRTUAL)
● API + SDK model: Simplifies agent integration for game companies and virtual world projects.
● Agent prompt interface: Coordinates the interaction between user input and the agent strategy engine.
● Strategic Planning Engine: Splits agent logic into high-level goal planning and low-level policy execution.
● Blockchain Integration: Potential on-chain wallet operators for decentralized agent governance.
GAME's architecture is highly customized for games or virtual environments, prioritizing real-time performance and continuous agent adaptation. While its role is not limited to the field of games, the system is clearly designed for virtual worlds and procedurally generated scenarios.
(3)Rig (ARC)
● Rust Workspace Structure: Separates functionality into multiple crates for clarity and modularity.
● Provider Abstraction Layer: Standardizes interactions with various LLM providers (OpenAI, Anthropic).
● Vector Store Integration: Supports multiple backends (MongoDB, Neo4j) for contextual retrieval.
● Agent System: Embeds retrieval augmentation generation (RAG) and dedicated tool use.
Rig's high-performance design benefits from Rust's concurrency model, making it ideal for enterprise environments that require strict resource management. Its conceptual clarity - through layered abstractions - provides strong reliability, but Rust's learning curve may limit the number of developers.
(4)ZerePy (ZEREBRO)
● Python-based: accessible to AI/ML developers familiar with Python codebases and workflows.
● Modular Zerebro backend: provides creative content generation, especially for social media and art.
● Agent autonomy: focuses on “creative output” such as meme, music, and NFT generation tasks.
● Social platform integration: includes built-in commands for Twitter-like functionality (post, reply, retweet).
ZerePy fills a gap for Python developers looking to deploy agents directly on social platforms. While ZerePy has a narrower scope than Eliza or Rig, its art- or entertainment-driven use cases thrive, especially in decentralized communities.
（1）Availability
● Eliza: Takes a balanced approach, with a moderate learning curve due to the complexity of multiple agents, but has a strong base of TypeScript developers.
● GAME: Designed for non-technical adopters in the gaming space, offering a no-code or low-code approach.
● Rig: More challenging; the Rust language strictly requires expertise, but it can reap high performance and reliability.
● ZerePy: Easiest for Python users, especially in creative or media-focused AI tasks.
(2) Scalability
● Eliza: The V2 iteration introduces a scalable message bus that improves concurrency, but multi-agent concurrency can be complex.
● GAME: Scalability is related to real-time game requirements and blockchain networks; performance will remain unchanged if game engine constraints are controlled.
● Rig: Naturally scalable through Rust's asynchronous runtime, suitable for high-throughput or enterprise-level workloads.
● ZerePy: Extensions are community driven and tested primarily in creative or social media environments, with less emphasis on large enterprise workloads.
(3) Adaptability
● Eliza: Highest adaptability to the plugin system, with broad model support and cross-platform integration.
● GAME: Specially adapted to the gaming environment, can be integrated into a variety of game engines, but is less suitable for other areas outside the gaming domain.
● Rig: Suitable for data-intensive or enterprise tasks; provides a flexible vendor layer for multi-LLM and vector storage.
● ZerePy: Targeted at creative output; easily extensible in the Python ecosystem, but with a narrow domain scope.
(4) Performance
● Eliza: Optimized for fast-moving social media or conversational tasks, performance depends on external model APIs.
● GAME: Real-time representation of game dynamics; success depends on the interplay of agent logic and blockchain overhead.
● Rig: High performance due to Rust's concurrency and memory safety, well suited for complex, large-scale AI processes.
● ZerePy: Performance depends on Python's speed and model calls; generally good enough for social/content tasks, but not targeted for enterprise-grade throughput.
All four frameworks have a combined market capitalization of $1.7 billion, and if the AI x Crypto industry follows the explosive growth pattern that has been seen in L1 blockchains, it has the potential to grow to more than $20 billion. For investors who believe that these frameworks (each serving a different market niche) will rise together under a broader "upward" trend, the market capitalization weighting approach may be the most prudent.
● Eliza (AI16Z): It is likely to continue to maintain the highest market share due to its established ecosystem, strong codebase, and upcoming enhanced version V2 (e.g., Coinbase Proxy Suite integration, TEE support).
● GAME (VIRTUAL): It is expected to gain further popularity in games/virtual worlds. Synergy with the VIRTUAL ecosystem ensures continued developer interest.
● Rig (ARC): May become a "hidden gem" for enterprise AI on Solana; as its partnership program matures, it can replicate the traction that other chain-specific frameworks have.
● ZerePy (ZEREBRO): Although its scope of application is small, it benefits from strong community momentum and the Python ecosystem, especially for creative and artistic use cases that are often overlooked by more general solutions.
(1) Technology Stack and Learning Curve
● Eliza (TypeScript) strikes a balance between accessibility and rich functionality.
● GAME provides an accessible API for games, but may target a niche group.
● Rig (Rust) maximizes performance at the expense of a higher complexity threshold.
● ZerePy (Python) is simple for creative applications, but lacks broader enterprise adoption.
(2) Community & Ecosystem
● Eliza: Top performer on GitHub, reflecting strong community engagement and broad applicability.
● GAME: Fast growth in games and virtual worlds thanks to VIRTUAL support.
● Rig: Targeting a small, skilled developer community, focused on high-performance use cases.
● ZerePy: A growing niche community built around creativity and decentralized art, whose development has benefited from its partnership with Eliza.
(3) Future Growth Catalysts
● Eliza: New plugin registry and TEE integration may further consolidate its leadership.
● GAME: Aggressively expanding through VIRTUAL’s ecosystem; accessible to non-technical users.
● Rig: Potential collaboration with Solana once developer traction increases, and focus on enterprise may lead to strong growth.
● ZerePy: Leverages Python’s popularity in the AI space and the cultural momentum around creative, community-driven projects.
The utilisation of machine learning has played a role in enabling Paul McCartney and Ringo Starr to transform an archival John Lennon demo into what’s likely the band’s last collaborative effort.
KikyoThe company is only the fourth entity that has been approved to offer crypto services to UK customers.
ClementAs convicted traffickers deliberated over the optimal offshore haven for their crypto wealth, US authorities listened in and pounced on the proceeds accrued through illicit darknet drug sales.
CatherineProShares CEO Michael Sapir stated that the new ETF was created specifically to help people short ETH.
ClementSingapore's cryptocurrency industry, once booming, faced a significant setback in 2022 due to a $2 trillion market crash and regulatory actions. However, the city-state remains focused on balancing innovation and investor protection in its evolving crypto landscape.
JoyThe jury took less than five hours to reach their guilty verdicts.
ClementInstead of requiring users to write complex queries with programming skills, individuals can now ask questions in plain language using Dune's natural language engine.
DavinThe security of Tellor's X account was compromised by a sophisticated phishing scam that enticed users with fraudulent airdrop offers.
KikyoBRICS nations unveils plans for their collective currency, shaping the future of global finance.
Hui XinThis collaboration has a clear objective: an Ethereum-based marketplace that stands as a testament to its dedication to preserving the rightful royalties of NFT creators upon each sale.
Catherine