Virtuals leads the way, CreatorBid stealthily advances, a glimpse into the latest developments in the AI Agent track.

Written by: @Defi0xJeff

Compiled by: zhouzhou, BlockBeats

Editor’s Note: The article evaluates the performance of several crypto AI projects in areas such as ecological construction, product iteration, community distribution, and token value. It concludes that Virtuals is the strongest in terms of speed and maintaining popularity, while CreatorBid, despite its slower execution, has a clear vision and focuses on the Bittensor smart agent ecosystem, showing potential for long-term growth. Overall, the 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 the sake of readability and understanding, the original content has been reorganized):

It has been about 7 months since the AI Agent craze began. This wave was initially started by the birth of @truth_terminal ➙ @pmarca invested in it ➙ someone issued tokens for it ➙ it began promoting the tokens ➙ @virtuals_io launched the agent tokenization platform ➙ AIDOL and conversational agents emerged ➙ alpha agent phase, @aixbt_agent rose ➙ framework phase, @elizaOS (formerly ai16z) initiated an 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 is basically a summary of the main stages of the AI Agent track.

Among the teams that have evolved from these stages, there are a few reliable AI agent teams - they remain active, continuously launching new products and features (although mainly relying on the trading fee income accumulated in the early stages to sustain themselves).

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.

Role of the Ecosystem Leader

These ecosystem leaders provide invaluable support:

• Has a powerful distribution network that can bring attention to your tokens and projects;
• Integration of products/services with the core of the ecosystem (i.e., aimed at potential users);
• Provide guidance and incubation services from 0 to 1 and then 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 central component of the crypto world - the community is key to whether tokens can form 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 various ecosystem leaders rise and fall. However, those projects that remain active stand out in the following areas:

• 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
• 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 is doing well, who is leading, who is lagging behind, and so on.

We will analyze from the following aspects:

• Products and Distribution
• AI / Intelligence Level
• Development Speed
• Token Value Capture Situation

Without further ado, let's first look at the first aspect:

Products and Distribution

In Web3, tokens themselves are often seen as a type of product. However, 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 toward a project, agents that trade or predict, with the goal of outperforming the market, and so on.

The success of a product 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 is using; what everyone cares about is whether its output is stable and 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, prediction agents, sports analysis, music, DeFi, and more.

Virtuals can be said to be the strongest in storytelling and shaping narratives, while also being the team that is best at listening to community feedback and iterating quickly (can be described as the "survivors").

However, although they offer a wide variety of services, there are actually only a few teams that provide products that deliver real value to users (rather than just entertainment).

Virtuals is the first pioneering player to launch an AI Agent startup platform, allowing anyone to publish conversational agents and bind a token. This mechanism is a double-edged sword—Virtuals can initially collect fees and gain value from these startups, but because anyone can publish, it has attracted a large number of short-term speculators and value harvesters who may repeatedly issue tokens or even run away immediately after launching.

Arc

Players like @arcdotfun have taken a completely different path.

They did not choose to build a "launchpad" and encourage as many projects as possible to go live, but instead focused on creating the AI Agent marketplace "Ryzome" by collaborating with a few high-quality projects to integrate their products and services into their MCP infrastructure.

In addition, they will launch a no-code / node-based agent building tool called "Ryzome Canvas", allowing users to connect to universal MCP server resources, as well as 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 none other than @elizaOS.

Eliza supports various integrations, such as secure execution through TEE, conducting transactions, analyzing real-time on-chain data, executing smart contracts, managing wallets, and more.

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 accomplish tasks (such as trading, social media automation, and business process automation).

For this reason, the user base of Eliza continues to grow, currently having 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 failed to seize the momentum and traffic dividend 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 the $ai16z liquidity pool), and promised to use part of the transaction fees to buy back $ai16z tokens.

However, so far, autodotfun has not shown any significant differences among its peers in the launch platform, and there haven't been any truly interesting or unique projects launched, which is somewhat disappointing.

AI / Intelligent Capability

As mentioned before, most of the time, the market is more focused on the "products" and "distribution" rather than the underlying architecture or the AI model itself.

But if you have a powerful and constantly 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 compared to a general model; a model trained on sports competition data, crowd intelligence, and real-time data will also have an advantage in predicting match results.

Bittensor is still the largest ecosystem with the most diverse intelligent models, and the only one truly dedicated to combining Bittensor subnet intelligence with AI Agent / Agentic workflows is @CreatorBid.

This team has performed poorly in distribution (slow new agent onboarding, slow iteration pace), but they have a clear goal in the direction of "steadfastly supporting Bittensor." (They have not officially announced it yet, but they may launch a subnet called SN98 Creator, further incentivizing the building of agentic workflows based on Creatorbid and going live.)

Development Speed / User Growth / Project Launch Rhythm

In Web3, if you are working on a long-term product, you must consider: how to keep the community engaged in the short to medium term.

If you cannot "entertain" the community, the token price tends to decline over time, as no one is willing to be stuck for the long term. In contrast, the market prefers projects that can continuously generate topics and build publicly.

Virtuals is the strongest player in this regard, openly developing, quickly fixing issues, actively listening to community feedback, and regularly launching new features or narratives to maintain users' ongoing interest, while also building their ACP. Additionally, they often have Genesis Launch events for new users to participate in.

Eliza's distribution capability ranks second, thanks to its developer network and partnerships with multiple L1/L2 collaborations. Eliza is also the preferred framework for deploying agents on other chains (non-Solana). autodotfun provides an easier onboarding path for the project.

Arc's Ryzome and Ryzome Canvas are in development. Once released, they could reignite interest in the ecosystem and potentially activate the launch of more Forge projects.

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 to launch its own subnet. The overall pace is relatively slow, hoping to speed up in the future.

Token Value Capture

$VIRTUAL is currently the strongest value-capturing token. It is the primary currency for LP construction 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 could be the second strongest. autodotfun has a daily trading volume of $2 million to $3 million (still far lower than Virtuals and other platforms), with some fees used to repurchase $ai16z. However, Eliza needs to quickly launch high-quality projects, especially those with a market cap over $10 million, otherwise, attention will still be focused on Virtuals.

The value capture of $arc comes from LP trading fees and future revenue streams generated by developers on Ryzome. However, this path is still in the early stages and requires time to materialize.

The token mechanism of $BID is the most unique, as its circulation is lower than similar projects, allowing for the release of tokens to incentivize platform activity. However, at present, these releases have not been well utilized, and trading volume remains low (ranging from $100,000 to $500,000 per day).

Summary

Each of the above items has its own advantages, but in the short to medium term, "distribution capability" + "ability to attract speculative funds" (i.e., trading volume) is the most essential moat.

The key to the system's operation is whether it can continuously generate excitement and attract players to keep betting in your "casino." In this regard, Virtuals is currently the best-performing project.

Whether they can maintain their popularity in the long term and transform it into real product capability 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 masses and truly commercialize agentic workflows.

Imagine this: an ever-evolving trading signal system that continuously outperforms the market, then transforms it 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 of the ecosystem, such as:

• @vana - Focus on data ownership
• @NousResearch - Reinforcement Learning
• @TheoriqAI - Liquidity Provision System
• @gizatechxyz——Focusing on finance / stablecoin-related agency

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 deeper decentralization of agent systems, and entrepreneurial opportunities at various layers of the technology stack.

Ultimately, the speculation heat may shift from individual agent tokens to the core infrastructure for building open AI systems. Perhaps we will see truly consumer-facing AI products that generate real income, rather than short-term speculative bubbles supported purely by "degens trading back and forth."