Dataline Raises $10.6M to Build a Data Network for AI Agents

Dataline, a startup building a decentralized data network for AI agents, has raised $10.6 million in seed funding and plans to hold a public token sale before the end of the year, according to reporting by The Block on October 21, 2026.

The round was co-led by Lattice and Samsung-backed Dispersion Capital, with participation from Symbolic Capital, Coinbase Ventures, Anagram, and Robinhood co-founder and CEO Vlad Tenev. Other backers include a former head of the CFTC and founders or partners from Venmo, Anchorage Digital, NEA, General Catalyst, Jump Capital, and FalconX, The Block reported. The financing closed in April through a simple agreement for future equity with a token warrant attached. The company declined to disclose its valuation, and no investor received a board seat, observer role, or advisory position.

Why agent data is becoming the bottleneck

AI agents can already browse, reason, and act, but reliable external data remains one of their weakest links. An agent that trades on a prediction market, manages a portfolio, or routes logistics needs more than a raw price feed. It needs to know whether the numbers it sees are trustworthy, fresh, and consistent across sources. Today that judgment is mostly hand-built by each developer, which slows down agentic products and makes them fragile.

That gap has been drawing capital all year. The market Dataline is targeting is one in which AI agents query on-chain and off-chain data and pay for it, a metered-data layer for the agent economy, as TradingView's coverage of the announcement noted. Dataline co-founder Nuo Shi framed the problem with a concrete market scenario, explaining that a stablecoin trading at $0.97 could be a buying opportunity or the early stage of a depeg, and that a simple price feed cannot distinguish between the two. Combining price data with peg-stability signals and cross-exchange anomaly information, Shi said, is what gives agents, bots, and smart contracts enough context to act.

The funding wave around agent infrastructure keeps widening. Recent weeks have seen raises for browser automation tooling aimed at agents, such as Reduck's $3 million pre-seed round, and security tooling for autonomous agents, as covered in recent reporting on the risks of dormant agents. Data reliability is the next layer in the stack.

How the Dataline network works

Dataline's platform is live with roughly 50 data sources connected, spanning on-chain and off-chain feeds. Its first "AI Data Launch" group includes Hyperliquid, Polymarket, Kalshi, Base, Robinhood Chain, BNB Chain, Sui, and TON. Developers building on Base and Robinhood Chain receive free access passes, a distribution move that puts the network in front of two of the most agent-active chains.

The core pitch is that Dataline does not just deliver numbers. Its own AI models score each data point for reliability, recency, and consistency, and attach the context an agent needs to make a judgment call. The company plans to expand the number of sources, upgrade its verification technology, and move into adjacent territory including robotics and AI training data. Dataline was founded in 2025 after winning a BNB Chain AI hackathon, and it plans to take a cut of data payments once it begins charging fees.

A token sale on Base, planned by year-end

The second half of the announcement is the token. Dataline plans to launch its token this year and run a public sale on Base using Uniswap's Continuous Clearing Auction system, with sale size and terms expected to be disclosed next week. The token is slated for network payments and data access. Data providers will stake tokens as a quality guarantee, and users will be able to lock up tokens to keep receiving data feeds instead of paying a monthly subscription, according to KuCoin's summary of the announcement.

That design puts Dataline squarely in a growing cohort of crypto-native agent infrastructure, protocols where the payment rails and the data rails are the same system. As Digital Today's coverage noted, the sale terms are not yet public, so market attention will shift to the size and valuation disclosed next week.

What to watch

Three things will determine whether Dataline matters. First, the auction terms. A fair, well-subscribed public sale would give the network a real user base; an overpriced one would leave it with speculators instead of agents. Second, verification quality. Scoring data reliability is an easy claim and a hard product. The network has to be meaningfully better than each team rolling its own checks. Third, distribution. Free passes on Base and Robinhood Chain are a start, but agent developers are the scarcest resource in this market, and winning them is a different contest than winning investors.

For now, the raise is another signal that capital is flowing toward the plumbing of the agent economy, not the models but the data, payments, and trust layers that let agents operate in the real world. Dataline has the funding, the investor list, and a live network. The token launch will show whether agents actually show up.