The Web3 x AI Market Map Series: Base

Web3 x AI Ecosystem Series: Base Ecosystem Market Map -– Abdul Al Ali, Venture Investor at Decasonic 

We are excited to kick off the Web3 x AI Ecosystem market map with Base, an ETH L2 developed by Coinbase and launched on August 9, 2023. Base is built on the OP stack and is EVM compatible. It leverages Coinbase’s 110m verified users (updated as of Q4 2022), and provides a distribution avenue for Coinbase users to go on-chain. Few ecosystems to date have done similar to Base in terms of launching, developing, and maintaining a large number of Web3 x AI projects that are rapidly gaining adoption and refining the emerging Web3 x AI landscape. This is in part, due to sheer resources and distribution channels available to Base ecosystem builders, fueled by the support of Coinbase and their active developments at the intersection of Web3 x AI. 

Coinbase has an extensive history in advocacy of the intersection of Web3 x AI, going back to 2018, when they welcomed their first VP of Data with the core vision of creating an open financial system by leveraging machine learning and AI. This is part of the reason why we are kicking off our ecosystem market map with Base, due to the dedicated commitment of Coinbase and the vision they had for the early innings of the advancing Web3 x AI supercycle. Moving to more recent times, this embrace by Coinbase extends to active developments by the Coinbase team, including AI tools: Coinbase’s Agentic Kit and the release of x402. 

The introduction of the ‘Based Agent,’ by @MurrLincoln kicked off key adoption of AI in the Base ecosystem, with a focus on advancements in AI Agents on Base capable of executing and interacting with other agents on-chain. This toolkit serves as an accessibility point for agents, allowing them to own and manage wallets. Their AI tools and integrations are often open-source, enabling builders to adopt their infrastructure and build on Base, onboarding more founders and projects. 

Toolkits on Base also tend to emphasize low-code or no-code platforms for agent creation and development, emphasizing accessibility - and enabling non-coders and non-technical founders to build on Base. Reflecting this advancement is Base being the preferred home to many specialized AI Agents as well, with multi-agent frameworks that focus on leveraging collaborative, specialized agents to solve stated goals. This includes the ACP framework, which is a core focus of Virtual Protocol and multi-agent frameworks, multi-agent coordination systems enabled by Autonolas.  

x402

The most recent agentic development by Coinbase is also their x402 payment protocol, which is an open-standard released on May 6, 2025. While released by Coinbase, this protocol is built on top of the Base ecosystem. It enables programmable payments, allowing developers and users to enable APIs, Apps, and AI Agent payments via the HTTP protocol. Agents can send an x402 request, and the server in turn responds with a HTTP 402 status code and payment details (including supported tokens like USDC). The server then verifies the on-chain payment, and returns the requested data and/or content. An example of the payment flow is highlighted below. 

Source: link.

This is a key advancement in the machine economy, enabling a new form of payment method and paving the way for composable machine-led service offerings. The distribution of this protocol is enhanced by key partners, including AWS, Circle, Anthropic and NEAR. 

Founder-led Advocacy

Beyond this, Jesse Pollak the co-founder of Base has been an advocate of using effective AI as a fundamental expectation for their team’s role, and leveraging AI to scale “impact materially quarter over quarter.” This builds on Tobi from Shoify’s ‘Reflexive AI usage is now a baseline expectation,’ article on X. This advocacy for team-led adoption of AI tools extends to ecosystem projects. Further, Pollak is a user of the protocols and frameworks that leverage AI on Base, including creating a digital twin through Freysa Protocol, named ‘jessexbt1.’ 

Coinbase AI Hackathon

The Coinbase AI hackathon (in partnership with former Replit team members) was a further key highlight for AI hackathons during Q1 ‘25, highlighting the commitment of Coinbase in advancing the AI ecosystem on Base. 

The State of Base

Owing to the success of Base in attracting key builders is the state of the Base ecosystem. The chain has successfully attracted $3.874B in TVL, with a growing number of daily transactions currently sitting at ~13M/transactions per day. Base further has a significant number of wallet addresses, currently at 170M+ addresses as of May 19, 2025. At Decasonic we believe that over the next few quarters and years, many more of those addresses will be owned by and operated by AI Agents, executing on-chain transactions and coordinating with other, specialized agents for transactions. 

Base Web3 x AI Market Map (2025)

Similar to our other market maps focused on Web3 x AI, we map the Base AI ecosystem across two axes:

• X-Axis – AI standard layers: Compute → Data → Model → Interface → Application.

• Y-Axis – Target audiences: Web3 Natives, Developers, Users, and Mainstream Audiences.

Note - as with our other market maps, this is a continued iteration and is a rapidly evolving, dynamic landscape. The current state of the ecosystem is unlikely to be the same over the next few quarters and years to come. To date, we have identified 174+ projects building on Base identified through our three pillars noted below.

The Three Pillars of Base Web3 x AI

Infrastructure

1. Model Layer 

   1. GenAI: Development, fine-tuning and distribution of GenAI Models. 

   2. Marketplace: Marketplace for publishing, discovering, buying, and selling AI Models.

   3. Agentic Framework: Infrastructure for the development and deployment of autonomous AI Agents, often fueled by underlying open-source models. 

      1. Projects operating at this intersection include Freysa AI, focused on providing frameworks for building AI “digital twins.” 

   4. Launchpad: Launchpad for the tokenization of AI models, enabling co-ownership of AI models launched off the platform. 

      1. An example of a project operating at this subcategory is Ora Protocol, which launched tokenized models through their IMO offerings. 

   5. Coordination: Collaborative model development and deployment

      1. An example of a project at this intersection is Synternet, which enables cross-chain intent for AI Agents.

   6. Inference: Enabling end users or developers to run inference on AI models, with an emphasis on model and output verifiability. 

2. Data Layer

   1. Marketplace: Marketplace for sourcing, curating, and trading datasets.

      1. Navigate, a data marketplace for AI Agents powers this intersection on Base. 

   2. Decentralized Training: Decentralized training process for AI models by distributing tasks across users, often specialized in nature. 

      1. Sapien operates in this category, building a decentralized data foundry. 

   3. Physical AI: Real-world data collection, annotation, and labelling for edge cases in robotics and Physical AI. 

      1. Auki, which operates a posemesh for Physical AI; primarily enabling machine perception and spatial computing. 

   4. Protocol: L1 or L2s optimized for data ownership, exchange, or validation

3. Compute Layer

Interface Layer

1. AI Agent Protocol: Protocols dedicated to creating, deploying, and managing AI Agents, often on-chain. 

   1. Morpheus is a project building a peer network for general purpose AI with a token on Base. 

2. AI Agent Launchpad: Launchpads specifically built to fund, deploy, and distribute AI Agents. 

   1. Virtuals Protocol is a key innovator and first mover at this intersection, kicking off the wave of AI Agents in Web3 during Q4 of last year. 

3. Analytics: Analytics platforms that leverage AI to generate insights, often incorporating real-time data parsing, summarization, and predictive intelligence.

   1. A key example is Kaito, which is advancing the InfoFi space powered by AI insights and predictive analytics. 

4. OS: Decentralized or AI-native operating systems designed to orchestrate AI agents, local models, or multi-modal inputs across devices or environments.

5. MCP: Decentralized Model Contextual Protocols (deMCPs) coordinating AI workloads or agents across distributed servers for collaboration or inference.

   1. Heurist, offers deMCP servers, with some of their recent MCPs including Cookie and AIXBT agents operating at this intersection. 

6. Physical AI: AI-enhanced physical devices such as robotics, sensors, or wearables. 

7. Chat: Interfaces that enable users to interact with large language models (LLMs) via chat-based prompts, often embedded in apps, dApps, or wallets.

8. Model: Platforms or apps allowing access to multiple foundation models or LLMs, often enabling model aggregation, comparison, or ensemble usage.

   1. Venice is an interface project offering access to AI models, primarily focusing on privacy and decentralization of access. 

9. GenAI: Platforms that provide access to generative AI models (text, image, audio, etc.), often with creative or productivity-oriented workflows.

10. AI Agents: Interfaces that enable users to interact with or customize autonomous AI Agents capable of acting on behalf of users or businesses.

11. DeFAI: Interfaces at the intersection of AI and DeFi that enable users to execute on-chain transactions using natural language or voice commands.

    1. Key projects at this subcategory include Bankr, which was recently embedded in Coinbase Wallet. 

Application Layer

1. Apps

   1. Dating: AI-powered applications that focus on enhancing dating experiences, often through personalization and AI-enhanced matchmaking.

2. SocialFi: Applications operating at the intersection of AI and SocialFi, enabling social interactions, creator economies, and on-chain reputation systems. Refer to our previously published market map on AI x SocialFi here: link.  

   1. Key projects operating at this intersection include Empyreal, which is launching their Bet on Bluff Prediction market in a few weeks. 

3. Vibecoding: Platforms that enable “vibe coding” and the low-code or no-code development of on-chain dApps using AI-assisted tooling and generation.

   1. On Base, Ohara is a key innovator at this intersection enabling users to create tokenized apps and games. 

4. AppStore: App stores dedicated to AI applications and agents. 

5. GameFi: Gaming applications at the intersection of AI and GameFi, often leveraging agents for gameplay, creation, or economic coordination. Refer to our previously published market map on the intersection of AI x GameFi here: link.  

   1. An example of this subcategory is Super Champs, which integrates AI Agents in their core gameplay. 

6. Agents

   1. Memes: AI Agent meme projects, often generating 24/7 content on X and other distribution avenues. 

   2. Music: AI Agents specialized in music generation, remixing, or collaboration. 

   3. Analytics: AI Agents providing analytics and insights, often focused on on-chain data, market trends, or DAO/community metrics.

      1. AIXBT, a one-time top Yapper on the Kaito leaderboard, provides on-chain analytics powered by its terminal. 

   4. Security: AI Agents focused on monitoring and improving security, often identifying risks in smart contracts, wallets, or protocol activity.

   5. Oracle: AI Agents that act as oracles, providing real-time data indexing, summaries, and feeds to smart contracts or other on-chain systems.

   6. Art: AI Agents that generate, curate, or collect digital art and NFTs, often participating in creative communities or on-chain auctions.

      1. Artto, an AI agent with a crypto wallet that collects NFTs, operates at this intersection. 

   7. DAOs: Autonomous DAOs coordinated or operated by AI Agents, enabling proposal creation, treasury actions, and on-chain governance execution.

      1. VaderAI, a key leading AI Agent powers this intersection. Though in more recent weeks, Vader has expanded beyond its core DAO focus. 

   8. Swarms: Collaborative networks of specialized AI Agents that function as a swarm to complete tasks, optimize workflows, or scale coordination.

   9. Health: AI Agents focused on providing personalized health insights, tracking data, or wellness suggestions based on user inputs or wearables. This could ultimately be integrated with Physical AI devices, providing users with  personalized agent(s) for their health insights. 

   10. Gaming: AI Agents tailored for gaming experiences, often acting as avatars, assistants, or creators aligned with eSports or gamer personas.

   11. NFTs: AI Agents focused on discovering, curating, or trading NFTs, often tied to digital identity, collections, or market sentiment analysis.

   12. DeSci: AI Agents that support decentralized science (DeSci), enabling research workflows, literature analysis, or experiment tracking.

   13. Companions: AI Agent-based companions or digital twins, often serving as persistent personal assistants or emotional/social support entities.

   14. Sports: AI Agents that analyze and provide insights into sports, often enabling fan engagement, fantasy leagues, or betting interfaces.

   15. Content: AI Agents focused on generating content across formats like video, blog posts, newsletters, or social media aligned with user goals.

Opportunities

Despite the strength of the Base AI ecosystem, we are still early. There are yet to be explored opportunities. Considering the current three pillars identified (Infrastructure, Application, and Interface) layers, 37.3% of the projects fall under Applications, 48.2% Interface, and 14.5% Infrastructure. 

Opportunities at the intersection of Base and Web3 x AI include: 

1. AI x RWA: Base is currently advancing the intersection of agentic, stablecoin payments primarily fueled by their x402 protocol; which enables agents to pay around the HTTP 402 status code. This unlocks multi-coordination, specialized agentic swarms centered around stablecoin adoption and growth, alongside other tokens that agents could pay with over time. Yet, this remains a relatively unexplored area in the Base AI ecosystem that remains to be rapidly advanced. 

2. Web3 Commerce: Building upon the x402 protocol, the opportunities at the intersection of Web3 x AI; fueled by open-source protocols similar to x402 can give rise to AI Commerce applications, dominated by AI Agents that can coordinate and transact on behalf of users. 

3. AI x SocialFi: Q4 ‘24 gave rise to the wave of AI chatbots launched off of Virtuals Protocol. Those bots were primarily social in nature, capable of interacting with users and generating 24/7 content. We envision the AI x SocialFi future as leveraging a mobile-distribution, product-first funnel that combines both specialized agent swarms with augmented human capabilities. This level of workflow will be present in social media applications, and will foster personalized content creation and enhanced user onboarding. 

4. Embedded DeFAI Interfaces: We imagine a future of multi-interface DeFAI applications directly embedded in Coinbase Wallet, fueled through the wallet app store and providing opportunities for users to be onboarded on-chain, funneled directly from Coinbase. Currently, DeFAI Interfaces like Bankr are currently available as an embedded application in Coinbase Wallet. 

   1. Building on this, Base has a high TVL in stablecoins and institutional players that could be used for on-chain activity. 

5. AI Native Apps: Out of the projects identified, we have identified two projects in the Apps subcategory of Applications, with one app being primarily dedicated to Dating. We see more applications being available on Base, fueled by the user funnel from Coinbase and the increasing commoditization of the infrastructure layer.

   1. These apps will enable co-ownership of output, and co-creation, empowering users to build and co-own AI applications and AI output. 

      1. Vibe coding is a category likely to fuel the number of AI native apps released on Coinbase. 

6. Coordination: Despite the presence of infrastructure on Base, and the emergence of AI Agent Swarms, this area is yet to note adoption in the Base ecosystem. 

7. Physical AI: This area remains underexplored in Web3 x AI in general, but Coinbase is well positioned considering their core distribution avenues and active investments. 

   1. Devices: This includes physical devices that are able to store user data on-chain, potentially similar to Limitless AI. 

   2. Robotics: Base could fuel the coordination mechanisms and data layer for Robotics, with more projects emerging similar to Auki. 

8. AI x GameFi: The gaming ecosystem on Base continues to evolve, and we see the advancements in AI Agent dominated and run games, with opportunities for gaming environments to serve as sandboxes for the development and refinement of both AI Agents and Physical AI Robotics. 

Conclusion

The Base AI ecosystem market map is rapidly evolving. Base benefits from a strong distribution funnel, primarily through Coinbase, Coinbase Wallet, and the rapid pace of open-source, AI-advancing innovations developed by the Coinbase team. At Decasonic, we believe the next phase in Web3 x AI adoption will be fueled by on-chain applications, utilizing the key underlying mechanisms of blockchain with a focus on co-ownership, co-ordination, and culture (what we refer to internally as the three ‘Cs’). 

Decasonic is a Web3 x AI venture and digital assets fund focused on advancing and working alongside founders and innovators at the intersection of Web3 x AI. If you are a founder, a potential founder, or partner investor - reach out to us at Decasonic, together, we can shape the future of this rapidly evolving frontier. 

The content of this material is strictly for informational and educational purposes only. It is not intended to constitute investment advice, nor should it be considered a recommendation or a solicitation to buy, sell, or hold any asset. Decasonic does not endorse investments in any specific tokens, and nothing in these blog posts should be construed as legal, tax, or financial advice. Please consult with a qualified professional advisor before making any financial decisions. Decasonic provides no warranties, whether expressed or implied, on the content provided in these blog posts, including its accuracy, completeness, or correctness. The opinions expressed here are those of the authors and do not necessarily reflect the views of Decasonic. Please note that Decasonic may hold a position in some of the tokens mentioned, including Virtuals. Decasonic is not liable for any errors or omissions in the content of this material or for any actions taken based on the information provided herein.