Weekly code commits are down roughly 75% from early 2025 highs and active developers have fallen 56%, even as overall GitHub usage accelerates. The gap is widening.
Developer activity in public crypto repositories has fallen to multi-year lows, with weekly code commits down approximately 75% from their early 2025 peak and the number of active developers declining 56% over the same period, according to data tracking open-source contributions across blockchain ecosystems.
The decline is playing out against a backdrop of surging overall GitHub activity, making the crypto-specific drop all the more conspicuous: the platform is gaining contributors in nearly every other technical domain, and artificial intelligence is absorbing the lion’s share of that growth.
The data from individual ecosystems tells a consistent story. Ethereum, which has approximately 1,759 public repositories and has historically been the most active developer ecosystem in crypto, has seen a 54% decline in activity over the past three months and a 34% drop in weekly active developers, who now number around 2,800. Other major chains are exhibiting similar trajectories. Most major blockchain networks are losing developer headcount, and while more experienced contributors now account for a higher share of commits — suggesting a consolidation rather than a complete collapse — the volume of new developers entering the space has dropped sharply.
The talent competition with artificial intelligence is the dominant structural explanation. LinkedIn’s January 2026 labor market report documents the creation of 1.3 million new AI jobs globally between 2023 and 2025. AI engineer positions expanded 13 times over that period; roles described as forward-deployed engineer and product manager grew 42 times. Crunchbase reports $211 billion in global AI funding in 2025, accounting for roughly half of all venture capital deployed worldwide. By comparison, PitchBook pegs crypto venture capital deal value at $19.7 billion over the same period — meaningful, but operating at a categorically different scale of capital attraction.
The talent drain has become visible in the departure of senior figures from some of the industry’s highest-profile projects in early 2026. Akshay BD, who spent five years building Solana’s developer ecosystem, posted a note saying he was ‘grateful to pass the torch.’ Nader Dabit left Eigen Labs to join Cognition, working on software agents that produce production-grade code autonomously. Kyle Samani stepped back from managing partner duties at Multicoin to explore AI and robotics. Anthony Rose left zkSync after four years. The clustering of these departures was not coordinated, but their concentration in the ecosystem-builder and infrastructure layer — roles that connect capital to projects and developers to distribution — was notable.
A separate factor is the shift toward closed-source development. As crypto projects mature and face greater competitive pressure, more teams are moving their codebases off GitHub and into private repositories — a logical business decision but one that makes the public commit data look worse than the true state of builder activity. Electric Capital’s most recent developer report, updated in January 2026, found that while the total number of monthly active developers fell roughly 7% year-over-year in 2024, developers with two or more years of experience hit an all-time high, up 27%. Core builders are staying; newcomers and short-term participants are leaving.
The long-run question the data raises is whether the structural decline in new developer inflows will ultimately affect the quality and pace of protocol innovation. Bear markets have historically seen developer attrition followed by recovery once prices stabilize. But the AI pull is qualitatively different from a price cycle: it is offering developers faster product cycles, larger immediate distribution, deeper capital pools, and roles at the center of what is widely described as the most important technological transition in a generation. Whether crypto’s promise of rebuilding financial infrastructure proves compelling enough to draw back — or retain — developers who have experienced what working at frontier AI companies feels like is the central talent question for the industry over the next two years.