Cryptopond
6 known investors
Pond (Cryptopond) is an agentic task marketplace where users post tasks or challenges and humans, AI agents, or robots compete to complete them, with payment only for verified results. It hosts diversified tasks such as lead generation, automation, and research contests.
Also known as CryptoPond Β· cryptopond.xyz Β· Pond
Investors Β· 6
Also in the syndicate Β· 3
Company profile
researched Aug 2026Pond, online at cryptopond.xyz, operates an AI task marketplace on which a requester describes a desired outcome and humans, AI agents, and β per the company β eventually robots compete to deliver it, with payment made only for accepted results. The site's workflow has three steps: a requester describes a task in natural language and Pond's AI model helps structure it, skilled people or agents pick it up, and the requester reviews submissions and pays for the ones that qualify. Tasks span lead generation, product feedback, bug testing, recruiting, content creation, and connecting third-party agents. Agents can be engaged individually, in collaboration as a team, or in competition where only the strongest submission is paid.
The company began in a different segment. According to a January 2025 profile, Pond was founded in 2023 and initially built a user search engine over on-chain data using a graph algorithm, then shifted toward crypto-native AI by building a graph network combining on-chain social, financial, and transactional data. Its stated model layer covered data infrastructure, model training, computation, and inference, targeting use cases such as predicting malicious addresses (a security model used by GoPlus), NFT and DeFi recommendations on platforms including Zora and Gearbox, and a dynamic DeFi fee model, with planned extensions into insider-trading detection and risk management. Its lead investor described Pond as building a foundation model on on-chain data using a graph neural network (GNN) architecture intended to predict relationships among accounts, nodes, and smart contracts, and to supply context to autonomous agents.
Public listings on the site show live task campaigns from third parties including the Ethereum Foundation (a "Deep Funding" contest with $5,000 in rewards), SELAT, Two Robots Studios, Pixero, and Alfi by Text.ai, alongside a directory of agents such as an AI lead-generation agent and an n8n workflow specialist.
Founding story
Pond was founded in 2023 β April 2023 per one profile β by Dylan Zhang (co-founder and CEO) and Bill Shi. Before founding Pond, Zhang was a founding member of Spanning Labs, an interoperability project backed by Galaxy Digital and other investors, where he handled business strategy, market research, and development; he also founded WednesDAO, an Ethereum Foundation-funded DAO connecting crypto founders.
Business model
Pond intermediates between task posters (startups and enterprises seeking outcomes such as leads, product feedback, bug reports, recruiting, or content) and a supply side of human contributors and AI agents. Requesters fund reward pools per task and pay only for submissions that pass verification; the site cites examples such as a customer paying for 35 of 99 submitted product-test reports and another paying 50 of 157 contributors who submitted screen recordings. The sources do not state Pond's own take rate or fee structure.
Not stated in the sources; the platform describes reward pools paid by task posters to contributors, with $31,715 paid out in total as of the site snapshot.
Traction
Site-reported platform metrics: 403,319 users, 189 countries covered, 20 agents at work, 39 tasks completed, and $31,715 paid out. Customer examples cite 245 registrants and 99 submissions in one week for Moatt (35 paid reports), 2,300 participants and 157 screen-recording submissions for PhotoBase on a $1,000 spend (50 paid), a 1,000-verified-lead task delivered overnight by 20 people and AI agents, and 83 participants running web tasks with TinyFish over four days at 8x the reward pool's capacity. Listed agents report usage figures such as 1,208 users/3,419 runs for an AI lead-generation agent. Pond's security model is reported as used by GoPlus.
Latest developments
April 2025: launch of Pond Agent Chat, a community-built, blockchain-data-trained copilot accessible via early access at cryptopond.xyz. The current website presents an agentic task marketplace with live third-party task campaigns, an agent directory, and cumulative platform metrics; it also lists individual backers including Dean Eigenmann (Project Blanc), Illia Polosukhin (Near), Alex Cheema (Exo Labs), Spencer Noon (OurNetwork), Tesa Ho (Flashbots), and Jeffrey Lau (ENS).
βΈFull profile β market position, technology, go-to-market, geography, history, risks & controversies
Market position
Positioned at the intersection of AI and crypto: originally as a decentralized AI model layer for on-chain data, and currently as an agentic task marketplace. The sources do not provide market-share or comparative ranking data.
Pay-only-for-verified-results task structure, an open contributor base combining humans and AI agents, and multiple agent delivery modes (individual, collaborative, competitive). Its earlier differentiation was a graph-neural-network model built for on-chain data rather than a general-purpose LLM.
Technology
Pond's earlier core technology was a graph-based AI model layer for blockchain data, using a graph neural network architecture to process graph-structured on-chain data and predict relationships and behaviors among wallets, accounts, and smart contracts, spanning data infrastructure, training, computation, and inference. The current platform uses an AI model to interpret free-form task descriptions and generate structured tasks, supports individual, collaborative, and competitive agent execution modes, and includes verification of submissions before payout. Pond Agent Chat, launched April 2025, is described as a community-built copilot trained on blockchain data for natural-language queries about on-chain behavior.
Go-to-market
Self-serve task creation on the website assisted by Pond's AI, plus branded task campaigns run with partner organizations and companies (e.g., Ethereum Foundation, TinyFish, Moatt, PhotoBase, Pixero) that bring their products or agents to Pond's contributor community. Earlier, the company ran "Frontier," a series of competitions to train web3 models, as an on-ramp to its model ecosystem.
Startups and enterprises that need discrete outcomes delivered β sales leads, product feedback and QA/bug testing, recruiting, content creation β as well as developers and agent builders seeking distribution and earnings for their agents. Earlier positioning targeted crypto developers and protocols needing on-chain AI models.
Geography
The site reports coverage of 189 countries. A domain-registry profile lists the company as United States-based.
History
Founded in 2023, Pond first built an on-chain user search engine using a graph algorithm, then expanded into crypto-native AI applications with a graph network over on-chain social, financial, and transactional data, offering models for security, recommendations, and DeFi. It raised a $7.5M seed round announced in November 2024, led by Archetype. In April 2025 it launched Pond Agent Chat, an AI copilot for exploring on-chain behavior. The current site presents Pond as a general agentic task marketplace for humans and AI agents rather than a crypto-only model layer.
Risks & controversies
The sources disclose no controversies. Reported scale is small in absolute terms (39 tasks completed and $31,715 paid out per the site), and the company's positioning has shifted materially from an on-chain AI model layer to a general task marketplace, which the sources do not explain.
Compiled by commissioned research from 8 cited public sources β announcements, filings, and press listed under research sources below.
Key figures
latest reportedCompany-reported or press-reported figures, each dated to when it was claimed β not independently audited.
Timeline Β· 4
launches, deals, and filingsPond launched Pond Agent Chat, an AI-native copilot for exploring on-chain behavior, built by the community and trained on blockchain data, available via early access at cryptopond.xyz.
The Ethereum Foundation posted a 'Deep Funding Contest - Level III' task on Pond with $5,000 in total rewards and 136 submissions.
$5K source β
TinyFish went live on Pond and handed its agent to the community; within four days 83 people were using it, with reported demand 8x the reward pool.
Archetype announced it led Pond's $7.5 million seed round; a later report lists participation from Cyber Fund, Delphi Ventures, Coinbase Ventures, Near Foundation and over 30 angel investors including Illia Polosukhin and Cynthia Wu.
$7.5M source β
Dated company events from announcements, filings, and press; legal rows summarize public dockets and regulator releases.
βΈResearch sources Β· 8
primary sources listed
- Cryptopondcryptopond.xyz Β· web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Cryptopond do?
- Pond (Cryptopond) runs an agentic task marketplace where people and AI agents compete on posted tasks and are paid for verified results.
- Who are Cryptopond's investors?
- Cryptopond's investors include Archetype, Coinbase Ventures, Web3com Ventures.

