Written by: @Defi0xJeff
Compiled by: zhouzhou, BlockBeats
Editor's Note: The article evaluates the performance of various crypto AI projects in ecosystem building, product iteration, community distribution, and token value, concluding that Virtuals is the strongest in speed and heat maintenance, while CreatorBid has a clear vision but executes slowly, focusing on the Bittensor smart agent ecosystem with potential for the long term. The overall AI agent track is still in its early stages, with future focus possibly shifting to infrastructure and real consumer scenarios.
Below is the original content (arranged for better readability):
About 7 months have passed since the AI Agent boom began. This wave was initially sparked by the birth of @truth_terminal ➙ @pmarca invested in it ➙ someone issued tokens for it ➙ it began promoting the token ➙ @virtuals_io launched the agent tokenization platform ➙ the AIDOL and conversational agent phase emerged ➙ alpha agent phase, @aixbt_agent rose ➙ framework phase, @elizaOS (formerly known as ai16z) initiated the open AI developer movement ➙ small-scale AI x gaming attempts (but no one survived) ➙ DeFAI phase (the vision remains strong, but execution is lacking)
This roughly summarizes the main stages of the AI Agent track.
From these stages, a few reliable AI agent teams have emerged — they remain active and continuously launch new products and features (although mainly sustained by early accumulated trading fee income).
Most importantly, some ecosystems remain strong, providing support for developers, helping product ideas start from zero, and promoting AI products and tokens from concept to successful launch.
Role of Ecosystem Leaders
These ecosystem leaders provide extremely valuable support:
Having a strong distribution network can bring attention to your tokens and projects;
Providing product/service integration with the core of the ecosystem (i.e., targeting potential users);
Providing guidance and incubation services from 0 to 1 and then to 10;
Support your ideas through investments and funding.
In the Web3 AI field, ecosystem leaders remain core pillars. Because the community is a core component of the crypto world — the community is key to whether a token can form a network effect (unlike traditional SaaS models that rely on subscription fees, Web3 projects rely on tokens to incentivize participation, accelerate growth, and user adoption).
In the past 7 months, we have seen several ecosystem leaders rise and fall. However, projects that remain active stand out in the following aspects:
Positioned as an AI Agent application store, developers/users can access services from Web2 and Web3 to enhance or automate their workflows — @arcdotfun
Build an economy for autonomous agents to trade with each other (and with humans) — @virtuals_io
Lead the largest Web3 open AI movement — @elizaOS
Combine Bittensor's subnet intelligence with AI Agent workflows, attracting more people to join the @opentensor (Bittensor) ecosystem — @creatorbid
This article will objectively analyze what each ecosystem does well, who is leading, who is lagging, etc.
We will analyze the following aspects:
Product and Distribution
AI / Intelligence Level
Development Speed
Token Value Capture Status
Without further ado, let's look at the first aspect:
Product and Distribution
In Web3, the token itself is often viewed as a product. But in this article, we define 'product' as goods or services that meet actual user needs.
In the Web3 AI field, most products revolve around 'financialization', meaning they are tools and intelligent services that help people make money — for example, alpha terminals, conversational agents that express sentiment about a project, agents that trade or predict, aiming to outperform the market, etc.
Whether a product is successful largely depends on 'distribution'. Generally speaking, this field is 90% distribution + 10% technical architecture. Few people in the industry care about what model your AI Agent uses; what everyone cares about is whether its output is stable, whether the insights and alpha it shares are genuinely useful.
Virtuals
@virtuals_io has the most diverse products within the ecosystem — including alpha signals, terminals, on-chain/off-chain data, agent workflows for auditing and security analysis, bots, investment DAOs, trading agents, predictive agents, sports analysis, music, DeFi, etc.
Virtuals can be said to be the strongest in storytelling and shaping narratives, while also being the best at listening to community feedback and iterating quickly (can be called 'survivor strong').
However, although the variety of services they provide is extensive, only a few teams are actually delivering products that provide real value to users (not just entertainment).
Virtuals is the first player to pioneeringly launch an AI Agent launch platform, allowing anyone to publish conversational agents and bind a token. This mechanism is a double-edged sword — Virtuals can collect fees and gain value from these launches in the early stages, but since anyone can publish, it attracts a lot of short-term speculators and value harvesters who may repeatedly issue tokens or even run away right after launch.
Arc
Players like @arcdotfun have taken a completely different path.
They did not choose to build a 'launch platform' and encourage as many projects as possible to go online; instead, they focused on creating the AI Agent market 'Ryzome', integrating these projects' products and services into their MCP infrastructure through cooperation with a few quality projects.
Additionally, they will launch a no-code/node-based Agent building tool called 'Ryzome Canvas', allowing users to connect to general MCP server resources and services and use cases provided by Arc partners to customize and create agent workflows (similar to Rayon Labs' Squad tool).
Users can sell these workflows or tokenize them and launch them through Arc's Forge (its launch platform).
Eliza
Among all frameworks, the most flexible and versatile is not @elizaOS.
Eliza supports various integrations, such as secure execution via TEE, conducting transactions, analyzing real-time on-chain data, executing smart contracts, and managing wallets.
This framework supports multi-agent systems, allowing developers to create a group of agents with different personalities, goals, and key performance indicators (KPIs) to collaboratively complete tasks (such as trading, social media automation, and business process automation).
Because of this, Eliza's user base continues to grow, currently with about 16,000 stars and 5,100 forks on GitHub.
However, although Eliza's framework is widely used, it initially lacked distribution channels. Unlike Virtuals, Eliza did not manage to capture the heat and traffic bonus in the early stages of AI Agent takeoff (the end of last year).
This situation changed a few weeks ago — Eliza launched @autodotfun, a launch platform priced in SOL (the next phase will introduce the $ai16z liquidity pool), and promised to use part of the trading fees to repurchase $ai16z tokens.
But so far, autodotfun has not shown significant differences among its peers, and there haven't been any truly interesting or unique projects launched, which is somewhat disappointing.
AI / Smart capabilities
As mentioned earlier, most of the time, the market is more concerned with 'products' and 'distribution' rather than the underlying architecture or the AI model itself.
But if you have a powerful and continuously evolving intelligent system, it is still possible to create more user-centric products.
For example: a model specifically trained on on-chain data will be stronger in analyzing on-chain information than a general model; a model trained on sports event data, crowd intelligence, and real-time data will also have an advantage in predicting match outcomes.
Bittensor remains the largest ecosystem with the most diverse intelligent models, and the only one truly committed to combining Bittensor subnet intelligence with AI Agent/Agentic workflows is @CreatorBid.
This team performs poorly in distribution (slow to launch new agents, slow iteration pace), but is clear about the direction of 'firmly supporting Bittensor'. (They haven't officially announced it yet, but may launch a subnet called SN98 Creator to further incentivize building agentic workflows based on Creatorbid and going online.)
Development speed / user growth / project launch rhythm
In Web3, if you are creating long-term products, you must consider: how to keep the community engaged in the short to medium term.
If you can't 'entertain' the community, the token price often declines over time because no one wants to be trapped for the long term. In contrast, the market prefers projects that can continuously generate topics and publicly build.
Virtuals is currently the strongest player in this area, openly developing, quickly fixing issues, actively listening to community feedback, and regularly launching new features or narratives to maintain ongoing user interest, while also building their ACP. They frequently have Genesis Launches for new users to participate.
Eliza's distribution capability ranks second, benefiting from its developer network and partnerships with multiple L1/L2s. Eliza is also the preferred framework for deploying agents on other chains (not Solana). autodotfun also provides an easier path to launch for projects.
Arc's Ryzome and Ryzome Canvas are in progress, and once released, they may boost ecosystem heat and activate more Forge project launches.
On the Creatorbid side, top agents recently launched new features (although the valuation range has not changed much). CB may be preparing to launch agents driven by the Bittensor subnet and launch its own subnet. The overall pace is slow, and we hope it can accelerate in the future.
Token Value Capture
$VIRTUAL is currently the token with the strongest value capture, serving as the primary currency for LP building in the Virtuals ecosystem, and agents entering Virtuals also need to use it. The recent Genesis Launch introduced Virgen points, which will flow to $VIRTUAL and other ecosystem tokens, further enhancing the holding value of $VIRTUAL.
$ai16z may be the second strongest. autodotfun has a daily trading volume of $2 million to $3 million (still far below Virtuals and other platforms), with part of the fees used to repurchase $ai16z. But Eliza needs to quickly launch quality projects, especially those with a market cap exceeding $10 million; otherwise, attention will still focus on Virtuals.
$arc's value capture comes from LP trading fees and the future revenue streams generated by developers on Ryzome. However, this path is still in its early stages and requires time to land.
$BID's token mechanism is the most unique because its circulation is lower than similar projects, allowing it to stimulate platform activity through token releases. However, currently, these releases have not been well utilized, and trading volume remains low (between $100,000 and $500,000 per day).
Summary
Each of the above projects has its own advantages, but in the medium to short term, 'distribution capability' + 'ability to attract speculative funds' (i.e., trading volume) are the most core moats.
Whether it can continuously generate heat and attract players to keep betting in your 'casino' is the key to system operation. In this regard, Virtuals is currently the best-performing project.
Whether they can maintain heat in the long term and translate it into real product strength is worth observing in the future.
Although @CreatorBid's execution still needs improvement, I personally have high hopes for them because their vision aligns with mine — bringing high-quality AI to the masses and truly commercializing agentic workflows.
Imagine: a continuously evolving trading signal system that consistently outperforms the market, then transforming it into a fully automated trading Agent — this is the vision of the SN8 Proprietary Trading Network.
It is still early in the market, and it is unclear who will ultimately win. More complex use cases are being handled by large teams outside the ecosystem, such as:
@vana - Focus on data ownership
@NousResearch - Reinforcement learning
@TheoriqAI - Create liquidity providing systems
@gizatechxyz - Focus on finance/stablecoin-related agents
In the future, how the leaders of the AI Agent ecosystem position themselves will determine whether they can seize the growth opportunities of the next cycle. We may also see more DeAI infrastructure landing, deeper decentralization of agent systems, and entrepreneurial opportunities at various layers of the tech stack.
Ultimately, speculative heat may shift from individual agent tokens to the core infrastructure of building open AI systems. Perhaps we will see truly consumer-facing AI products that generate real income, rather than purely relying on short-term speculative bubbles supported by 'degens trading back and forth'.
