TLDR:
- Uniswap API won 52.4% of 554,137 MetaMask swaps over 99 days, beating all rivals combined by count
- The API recorded the lowest median slippage of 0.21–0.88 bps and a 0.12% failure rate across all providers
- OKX’s volume lead of 25.3% stemmed from whale concentration, with one signer routing $42.6M in six trades
- Excluding top 100 wallets per provider, Uniswap led adjusted volume at 32.9%, ahead of Kyber and 1inch v6
The Uniswap API has emerged as the dominant routing provider inside MetaMask’s swap architecture on Ethereum mainnet.
An independent researcher identified as Vaish analyzed 554,137 successful swaps over 99 days, totaling $567.8 million in volume.
The study found the Uniswap API won 52.4% of all routed transactions by count. That figure exceeds every competing provider combined, raising questions about how AMM-based routing holds its ground against RFQ desks.
Uniswap API Leads on Transaction Count and Execution Quality
MetaMask operates a multi-provider swap architecture where each trade request goes out to roughly a dozen routing providers simultaneously.
The platform selects the best quote net of fees and presents it to the user. Since Uniswap’s API integration in March 2026, it has consistently won more of those head-to-head quote competitions than any rival.
Realized execution data backs the routing outcome. Vaish’s study measured slippage as the deviation between a user’s actual fill and the on-chain mid-price at execution time.
The Uniswap API posted the lowest median slippage across all five trade-size buckets, ranging from 0.21 to 0.88 basis points. Competitors recorded medians between 1 and 27 basis points.
Vaish summarized the finding directly: “Among the hundreds of thousands of ordinary users whose wallet shops their order across a dozen providers, Uniswap is the routing outcome more often than every competitor combined, and the settled trades justify the routing.”
The Uniswap API also recorded a failure rate of just 0.12%, the lowest among all major providers. Competing platforms ran failure rates between 0.27% and 0.64%.
At MetaMask’s transaction volume, that gap translates to thousands of avoided failed swaps and their associated gas costs.
The routing advantage held across pair types, gas conditions, and times of day, with transaction share staying between 52% and 55% in every UTC hour.
OKX Volume Lead Tied to Whale Concentration, Not Broad User Demand
By raw dollar volume, OKX led the dataset at 25.3%, compared to Uniswap’s 21.3%. However, the researcher found that figure came from extreme wallet concentration. OKX’s $143.8 million in volume originated from just 13,850 wallets, with the top 10 accounting for 48% of it.
A single intermediary contract operated by one signer sent $42.6 million through OKX across just six transactions. That entity alone represented roughly 30% of OKX’s total volume in the study period.
On this distinction, Vaish was direct: “Routing share by transaction count is a tally of independent quote competitions, one per user decision. Routing share by raw volume is that same tally reweighted by a few large actors.”
Uniswap, by contrast, drew its volume from 134,876 wallets, with only 5.4% concentrated among its top 10. After excluding the top 100 wallets from each provider, Uniswap led the adjusted volume ranking at 32.9%. Kyber followed at 18.2%, with 1inch v6 at 15.7%, OKX at 13.3%, and 0x at 12.9%.
On routes involving five or more legs, OKX captured 43.1% of volume while Uniswap dropped to 4.1%. That split reflects a structural divide: AMM-based routing performs strongest on standard retail trades, while RFQ desks hold an edge on complex, large-ticket routes above $100,000.
The Uniswap API was the only major provider whose share rose consistently under progressive whale exclusion.
