Original title: AI Agents: Ecosystem Leaders Original author: @Defi0xJeff Original translation: zhouzhou, BlockBeats
Editor's note: The article evaluates the performance of multiple crypto AI projects in ecosystem building, product iteration, community distribution, and token value. It finds that Virtuals is the strongest in speed and maintaining heat, while CreatorBid, despite slow execution, has a clear vision and focuses on the Bittensor intelligent agent ecosystem, with long-term potential to look forward to. The overall AI agent track is still in its early stages, and future focus may shift towards infrastructure and real consumption scenarios.
The following is the original content (for ease of reading, the original content has been slightly edited):
About 7 months have passed since the AI Agent craze began. This wave initially started with the birth of @truth_terminal ➙ @pmarca invested in it ➙ someone issued tokens for it ➙ it began promoting the tokens ➙ @virtuals_io launched an agent tokenization platform ➙ AIDOL and conversational agents phase emerged ➙ alpha agent phase, @aixbt_agent rose ➙ framework phase, @elizaOS (formerly ai16z) initiated the open AI developer movement ➙ small-scale AI x gaming attempts (but no survivors) ➙ DeFAI phase (the vision remains strong, but execution is lacking)
This is roughly the main stage summary of the AI Agent track.
From these stages of evolution, there are a few reliable AI agent teams - they are still active, constantly launching new products and features (although mainly relying on transaction fee income accumulated in the early stages).
Most importantly, there are still some ecosystems that remain strong, providing support to developers, helping product ideas start from scratch, and driving AI products and tokens from concept to successful launch.
The role of ecosystem leaders
These ecosystem leaders provide extremely valuable support:
· Has a strong distribution network that can bring attention to your tokens and projects;
· Provide product/service integration with the core of the ecosystem (i.e., targeting potential users);
· Provide guidance and incubation services from 0 to 1 to 10;
· Support your ideas through investment and funding.
In the Web3 AI field, ecosystem leaders remain a core pillar. Because the community is a core component of the crypto world - the community is key to whether a token can achieve network effects (unlike traditional SaaS models that rely on subscription fees, Web3 projects depend on tokens to incentivize participation, accelerate growth, and user adoption).
In the past 7 months, we have seen multiple ecosystem leaders rise and fall. But those projects that remain active stand out in the following aspects:
· Positioned as an app store for AI Agents, developers/users can access Web2 and Web3 services to enhance or automate their workflows - @arcdotfun
· Build an economy where autonomous agents trade with each other (and with humans) - @virtuals_io
· Leading the largest Web3 open AI movement - @elizaOS
· Combine Bittensor's subnet intelligence with AI Agent workflows to attract 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 from the following aspects:
· Product and distribution
· AI / Intelligence Level
· Development speed
· Token value capture situation
Without further ado, let's first look at the first aspect:
Product and distribution
In Web3, tokens themselves are often also seen as a product. But in this article, we define 'products' 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 - such as Alpha terminals, conversational agents that express sentiment towards certain projects, agents that trade or predict outcomes, 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. Very few in the industry care about what model your AI Agent uses; everyone is more concerned about whether its output is stable and whether the insights and alpha it shares are genuinely useful.
Virtuals
@virtuals_io has the most diverse products in the ecosystem - including alpha signals, terminals, on-chain/off-chain data, agent workflows for auditing and security analysis, robots, 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 (they can be called 'survivors').
However, while they offer a wide variety of services, there are actually only a few teams that truly provide real value to users (rather than 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 initially could collect fees from these launches and gain value, but because anyone can publish, it attracted a large number of short-term speculators and value harvesters, who might repeatedly issue tokens and even run away right after going live.
(However, Virtuals is developing ACP, hoping we can soon see some flagship agent products and services)
Arc
Players like @arcdotfun have taken a completely different path.
They did not choose to build a 'launch platform' and encourage as many projects to go live as possible, but instead focused on creating an AI Agent market 'Ryzome', integrating the products and services of these projects into their MCP infrastructure through collaboration with a few quality projects.
In addition, they will also launch a no-code/node-based Agent building tool called 'Ryzome Canvas', where users can access general MCP server resources and services and use cases provided by Arc partners to custom create agent workflows (similar to Rayon Labs' Squad tool). Users can sell these workflows, or tokenize them, and go live through Arc's Forge (its launch platform).
(In short, Arc is taking the 'polish the product first, then talk about distribution' route. Ryzome will soon open for testing.)
Eliza
Among all frameworks, the most flexible and variable is not @elizaOS.
Eliza supports various integrations, such as secure execution via TEE, trading, analyzing real-time on-chain data, executing smart contracts, managing wallets, etc.
The 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, business process automation).
As a result, Eliza's user base continues to grow, currently with about 16,000 stars and 5,100 forks on GitHub.
However, while Eliza's framework is widely used, it initially lacked distribution channels. Unlike Virtuals, Eliza did not manage to capture the heat and traffic bonus during the early stages of AI Agent's rise (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 a $ai16z liquidity pool) and promised to use part of the transaction fees to repurchase $ai16z tokens.
But so far, autodotfun has not shown significant differentiation in similar launch platforms, nor have any genuinely interesting or unique projects been launched, which is somewhat disappointing.
(Eliza's biggest advantage and disadvantage actually lie with @shawmakesmagic: without Shaw's countless high-intensity investments, this framework would not exist; but he also often 'crashes' and makes some questionable decisions, leading to market FUD, a situation that has occurred multiple times.)
AI / Intelligent Capability
As mentioned earlier, most of the time, the market is more focused on 'products' and 'distribution', rather than the underlying architecture or the AI model itself.
But if you have a strong and constantly evolving intelligent system, it is still possible to create more user-centered 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 game outcomes.
Bittensor remains the largest ecosystem with the most diverse intelligent models, and the only one truly dedicated to combining Bittensor subnet intelligence with AI Agents / agentic workflows is @CreatorBid.
This team underperformed in distribution (slow to launch new agents, slow iteration pace), but has a clear goal in 'steadfastly supporting Bittensor'. (They haven't officially announced it yet, but they may launch a subnet called SN98 Creator to further incentivize building agentic workflows based on Creatorbid and going live.)
Development speed / user growth / project launch pace
In Web3, if you are building a long-term product, you must think about: how to keep the community engaged in the short to medium term.
If you can't 'entertain' the community, the token price will often decline over time because no one wants to be stuck long-term. In contrast, the market prefers projects that can continuously generate topics and build publicly.
Virtuals is the strongest player in this regard, developing openly, quickly fixing problems, actively listening to community feedback, and regularly launching new features or narratives to maintain ongoing user interest, while also building their ACP. In addition, they often have Genesis Launch to engage new users.
Eliza's distribution capability ranks second, thanks to its developer network and partnerships with multiple L1 / L2. Eliza is also the preferred framework when deploying agents on other chains (non-Solana). autodotfun also provides a more streamlined on-ramp for projects.
Arc's Ryzome and Ryzome Canvas are progressing, and once released, they may drive the ecosystem's heat back up and potentially activate more Forge project releases.
On the Creatorbid side, top agents have recently launched new features (although the valuation range has not changed much). CB may be preparing to launch an agent driven by the Bittensor subnet and launch its own subnet. The overall pace is slow, hoping to speed up in the future.
Token value capture
$VIRTUAL is currently the strongest token in value capture; it is the main currency built from LP 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 daily trading volumes of $2 million to $3 million (still far below Virtuals and other platforms), with some fees used to repurchase $ai16z. However, Eliza needs to quickly launch quality projects, especially those with market caps exceeding ten million dollars, otherwise attention will remain focused on Virtuals.
$arc's value capture comes from LP transaction fees and future revenue streams generated by developers on Ryzome. However, this path is still in the early stages and needs time to land.
$BID's token mechanism is the most unique because its circulation is lower than that of similar projects, which can incentivize platform activity through token releases. However, so far, these releases have not been well utilized, and trading volume remains low (between $100,000 and $500,000 daily).
Summary:
Each of the above projects has its advantages, but in the short to medium term, 'distribution capability' + 'ability to attract speculative funds' (i.e., trading volume) is the most core moat.
Whether they can continuously generate heat and attract players to keep betting in your 'casino' is the key to the system's operation. In this regard, Virtuals is currently the best-performing project.
Whether they can maintain heat in the long term and convert 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 - to bring high-quality AI to the public and truly commercialize agentic workflows.
Imagine: a constantly evolving trading signal system that continuously outperforms the market, which is then transformed into a fully automated trading Agent - this is the vision of the SN8 Proprietary Trading Network.
We are still in the early stages of the market, and it is unclear who will ultimately prevail. More complex use cases are being handled by large teams outside the ecosystem, such as:
· @vana - Focused on data ownership
· @NousResearch - Reinforcement learning
· @TheoriqAI - Building a liquidity provision system
· @gizatechxyz - Focused 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 being implemented, a deepening of the decentralization of agent systems, and entrepreneurial opportunities at various layers of the tech stack.
Ultimately, speculative enthusiasm may shift from individual agent tokens to the core infrastructure that builds open AI systems. Perhaps we will see truly consumer-facing AI products that generate real income, rather than relying purely on 'degens trading back and forth' to support short-term speculative bubbles.
"Original link"
