AI-agent Platforms: Comparative Analysis

Introduction

In our December research on tokenized AI agents and the infrastructure around them, we examined the leading projects of that time (yes, in our industry, events from just a month ago are considered "that time"). Among other things, we took an in-depth look at a couple of platforms for launching AI agents: Virtuals Protocol and Vvaifu.fun. At the beginning of this year, the narrative of AI agents has logically expanded, and the narrative of platforms for launching AI agents has evolved into a successful sub-narrative. Even projects that have not yet launched on mainnet, such as Hey Anon and AI Rig Complex, are already valued at several hundred million dollars! In today’s study, we will delve deeper into this segment of the AI market, explore its main players, and attempt to identify potential winners through comparative analysis.

Important Notes and Disclaimers

Unlike our previous comparative analysis of projects from leading ecosystems, in this case, we cannot rely on quantitative analysis. The projects under review are, firstly, too young, and secondly, the market valuation of AI projects is not currently based on specific metrics. This is, in fact, the primary advantage of this narrative – it is extremely challenging to evaluate, and therefore, its upside in a bull market is virtually unlimited. Consequently, we cannot apply objective quantitative analysis but are compelled to adopt a largely subjective analysis of the products. Within this subjective evaluation, we will strive to remain as impartial as possible since we are neither token holders nor ambassadors of any project.

Comparative Analysis of AI-Agent Platforms

Analysis Criteria

Here are the four key criteria we used for our analysis:

Criterion #1: Accessibility & UX

We focused on minimizing barriers for potential users. Platforms should impose as few restrictions as possible, operate on popular blockchains with strong network effects, and support integration with widely-used wallet interfaces.

Criterion #2: Agent Customization Flexibility

Creating an AI agent should be as simple as possible while offering the flexibility to cater to users’ most complex customization needs. Agents should also support as many popular social platforms as possible, such as Twitter (X), Discord, and Telegram. Additional benefits include the ability to create agents without requiring a token or link agents to existing tokens.

Criterion #3: Market Success of Agents

This is the sole quantitative criterion in our analysis. We calculated the combined market capitalization of the five most successful agents launched on a platform (excluding the platform’s "native" agent). This provides insight into how well the market receives the tokens of agents created on each platform.

Criterion #4: Project Tokenomics

Here, we focus not on the token distribution among holder groups (which can be reviewed separately) but on the applicability and utility of the platform’s native token. Key considerations include its purpose and whether the project’s success is likely to increase the token’s value.

Virtuals Protocol

The market capitalization of the native token $VIRTUAL at the time of writing is $3.5 billion.

Description

We previously explored Virtuals Protocol in depth in our December research.

In brief, this project is a trailblazer in the AI Agent Platforms sector, positioning it as the leading representative of the narrative (as of the time of writing).

Accessibility & UX

The protocol operates on the Base blockchain, which is the second most popular among crypto users. It has not been reported to impose restrictions or bans on users and supports a wide range of compatible wallets. However, to achieve a perfect score in this category, the platform would need to integrate with the leading blockchain Solana and popular wallets like Phantom. Score for this criterion – 2 points

Agent Customization Flexibility

The agent generation system employed by Virtuals Protocol offers limited flexibility and customization. The primary customization available when defining an agent’s "personality" is their "biography" which, while allowing unlimited text input, lacks the depth needed for broader customization.

Market Success of Agents

Top-5 tokens launched on the platform (excluding $toshi, whose token existed long before the agent was created on Virtuals):

• $aixbt: $665 million
• $game: $200 million
• $vader: $90 million
• $rekt: $80 million
• $luna: $75 million

Total Market Capitalization: $1.1 billion. Based on this criterion, the project is the undisputed leader – 5 points.

Project Tokenomics

We have previously described the role of the token within the Virtuals ecosystem. In summary, the platform’s tokenomics are designed so that the success of AI agents launched on the platform directly drives the token’s price growth. Moreover, we can already observe the positive impact of this mechanism. In our view, the tokenomics of Virtuals are exemplary – 5 points.

Vvaifu.fun

The market capitalization of the native token $VVAIFU at the time of writing is $65 million.

Project Description

Vvaifu.fun was also reviewed in our December research, and little has changed since then. It can be considered the first such project on the Solana blockchain, successfully launching several tokenized agents.

Accessibility and UX

The platform is deployed on the Solana blockchain and supports the most popular wallet among users – Phantom, which is a significant advantage. From a user interface perspective, the platform offers an intuitive design for launching and trading agents. Our score for this criterion – 4 points.

Agent Customization Flexibility

The process of creating an agent on the platform is, on the one hand, highly customizable but, on the other hand, somewhat demanding. Users must fill in a significant number of fields to define various agent characteristics. Additionally, at the beginning, the user must select the agent’s framework (native Dasha, Eliza by ai16z, or ZerePy by Zerebro). In our subjective opinion, this provides experienced users with more freedom but might intimidate casual users unwilling to navigate the nuances of such choices.

As of the time of this research, agents created on vvaifu.fun can communicate on Twitter (X) and Telegram. The project team has promised to add support for Discord by the end of January. Score for agent customization flexibility – 3 points.

Agent Market Success

Top-5 tokens launched on the platform (excluding the native $vvaifu token and $usa, which existed before the platform’s launch):

• $aimoca: $47M
• $build: $34M
• $stoic: $14M
• $soltoshi: $8.5M
• $lios: $7.5M

Total market capitalization: $111M. Score for this criterion – 4 points.

Tokenomics

The tokenomics of vvaifu.fun are straightforward and effective regarding token utility and application. The protocol employs a deflationary model: the native token is burned each time an agent is launched through the native framework or when creators unlock additional features for their agents. As a result, the project’s token price is directly tied to the growing interest in the platform. Score for tokenomics – 4 points.

Top Hat

The market capitalization of the native token $HAT at the time of writing is $85 million.

Project Description

Top Hat is a platform for creating and trading AI agents on the Solana blockchain, closely resembling pump.fun in both design and functionality.

Accessibility and UX

The platform operates on Solana, the most popular blockchain among retail users, with no restrictions and full compatibility with the Phantom wallet. However, it does not support EVM blockchains or EVM-compatible wallets, which can be seen as a limitation.

As mentioned earlier, the platform’s UX closely mirrors pump.fun, though it appears simpler in some respects. The interface is intuitive and user-friendly, making it accessible for all levels of users. Score for this criterion – 4 points.

Agent Customization Flexibility

The process of creating an AI agent on the platform is similar to launching a token on pump.fun – it involves just one step. Despite its simplicity, it provides a decent level of customization, requiring the user to fill out fields such as Name, Description, Personality, and Instruction.

Agent Customization Flexibility

Additionally, there is an optional Knowledge Base field, where users can input a large amount of text. This adds to the platform’s customization depth. Overall score for this criterion – 4 points.

Market Success of Agents

Top-5 tokens launched on the platform (excluding the native $HAT token):

• $kolin: $33M (a token linked to multiple agents)
• $rosscoin: $26M
• $uai: $17M
• $kwantxbt: $14M
• $aaahhmm: $9M

Total market capitalization: $99M. The project ranks 3rd place by this parameter – 3 points.

Project Tokenomics

There is no information on the platform’s website or documentation regarding the utility or purpose of the native token $HAT. As of now, it can only be regarded as the platform’s official meme token. While this isn’t necessarily a negative for the token itself (meme tokens have proven popular in the market), it leads us to assign the minimum score for this criterion – 1 point.

Holoworld

The market capitalization of the native token $AVA at the time of writing is $230 million.

Project Description

Holoworld is a platform for creating tokenized AI agents on the Solana blockchain. The platform stands out for its high degree of customization for created agents, referred to as "virtual avatars".

Accessibility and UX

The platform operates on Solana but also allows the connection of EVM-compatible wallets for agent creation (excluding trading functionality). This is a highly advantageous feature, as it helps attract users from the EVM ecosystem.

The platform’s user interface is well-designed, intuitive, and aligns perfectly with the product’s purpose, posing no difficulty for users. Score for this criterion – 5 points.

Agent Customization Flexibility

The agent creation process is the platform’s main competitive advantage. Agent customization is extensive, with the creation page divided into four sections: Basic, Personality, Knowledge, and Avatar. Notably, most fields, apart from the essential ones, are optional.

Agent Customization Flexibility

The platform provides users with significant freedom to bring their creative visions to life without forcing them to create overly specific agents. While, at present, agents created on the platform only support Twitter (X) as their primary platform, the customization flexibility earns a score of 5 points.

Market Success of Agents

Top-5 tokens launched on the platform (excluding the native $AVA token):

• $digimon = $26M
• $spore = $25M
• $dwain = $12M
• $fme = $2.5M
• $ian = $0.5M

Total market capitalization = $66M, placing it in 4th position. Score – 2 points.

Tokenomics

The tokenomics of Holoworld is straightforward and similar to that of vvaifu.fun: to launch an agent on the platform, users must spend “credits.” The only way to acquire credits is by burning $AVA tokens. According to Solscan, only about 0.005% of the total token supply has been burned so far. Additionally, there are unconfirmed reports of a future staking mechanism for $AVA tokens. Final tokenomics score – 3 points.

Spectral

The market capitalization of the native token $SPEC at the time of writing is $120 million (with FDV at $880M).

Project Description

Spectral is a protocol for creating tokenized AI agents, with the unique feature that all agents maintain trading accounts on the decentralized exchange Hyperliquid, and all interactions occur via chat directly on the platform. The tokens of these agents are traded on the Base blockchain.

Accessibility and UX

The platform operates on the Base blockchain, supporting key EVM-compatible wallets like Metamask and Rabby. It features an extremely minimalist user interface. Score for this criterion – 3 points.

Agent Customization Flexibility

The platform cannot boast high flexibility in agent customization. However, the creation process is innovative, conducted through a chat-bot interaction. Score for this criterion – 1 point.

While it might seem unfair to penalize the platform for the lack of cross-platform presence (as it was never its intention), we struggle to understand how users would discover these agents without directly visiting the platform. This lack of market presence is a significant drawback. Score for this criterion – 1 point.

Token Success

Top-5 tokens launched on the platform:

• $spectr = $3.8M
• $rhj = $0.5M
• $cya = $0.2M
• $specy = $0.1M
• $yuki = $0.1M

Total market capitalization = $4.7M, 5th place – 1 point.

Project Tokenomics

The tokenomics of Spectral differ significantly from other analyzed projects. First, the native token $SPEC serves as a governance token, a utility more common in DeFi projects. Second, Spectral is the only project among the five analyzed to implement a vesting schedule, with only 10% of the total supply currently circulating.

Despite the token’s utility, we assign a low score for tokenomics due to limited applicability – 2 points.

Final Results Table

Below is the final table of the analyzed projects, showcasing their market capitalization, scores across our evaluation criteria, and the total score:

Conclusion

Based on the analysis conducted in this study, Holoworld ($AVA) emerges as the most investment-attractive project, as it scored equally to Virtuals Protocol ($VIRTUAL) but has a significantly lower market capitalization. However, it is essential to remember that our analysis is based on basic logic and publicly available information; we do not possess insider knowledge that could critically impact token valuations, either positively or negatively. Moreover, our criteria were intentionally limited, omitting factors such as exchange listings, liquidity depth on DEX pairs, and other nuances. We strongly recommend conducting further independent research to develop your own investment thesis, which may differ from ours.

Additionally, it is worth noting that tokens for unreleased projects, such as AI Rig Complex and Hey Anon, are already being traded. These projects are valued highly by the market (over $300M for the former and $200M+ for the latter). When these projects go live, they could lead to a significant revaluation—not only of themselves but also of their competitors. Our analysis methodology might prove useful for assessing these and other new entrants in the AI-agent platform segment.

Exercise caution in your investment decisions and be mindful of the risks inherent in this nascent AI sector, especially within the even younger segment of AI-agent platforms. But as the saying goes, where there are high risks, there are also high rewards. Best of luck!