Crypto rug pulls have decreased by 66% in 2025 compared to 2024, but losses have surged to nearly $6 billion. The situation is driven primarily by sophisticated scams in memecoins, most notably the collapse of Mantra’s OM token. Although awareness and tools to combat these frauds are improving, the risk remains significant as bad actors adapt their tactics.
In 2025, the number of crypto rug pulls has dropped significantly by 66% compared to the previous year. Data from DappRadar indicates that while only seven rug pulls have occurred this year, the financial impact has increased drastically, with losses nearing $6 billion. Notably, 92% of these losses stemmed from the collapse of Mantra’s OM token, which its founders deny was a rug pull.
DappRadar’s analyst, Sara Gherghelas, states that rug pulls are becoming less frequent yet more damaging. The scams have become increasingly sophisticated, often supported by teams with professional branding and well-thought-out narratives, making them harder to detect.
The characteristics of rug pulls have shifted; initially predominant in DeFi protocols and NFT projects, the current trend indicates that memecoins are now the primary vehicles for these scams. A notorious example is the Libertad project’s Solana token, LIBRA, which surged to a market cap of $4.56 billion following a post by Argentina’s president but subsequently plummeted by over 94% after the post was deleted.
Gherghelas warns that rug pulls and exit scams continue to pose a significant threat, especially where projects can quickly surge in popularity before vanishing with investor funds. Indicators of potential scams include unexpected rises in unique active wallets, high trading volumes with low user engagement, and projects lacking transparent smart contracts or developer identities.
As the cryptocurrency industry evolves, so too do the strategies employed by scammers. However, improvements in user tools for identifying these threats are emerging. Gherghelas notes that while eradicating rug pulls entirely may be unrealistic, better user education can substantially mitigate their effects.
