Figment’s institutional Ethereum staking versus Consensys and Kiln

Figment holds the position of largest non-custodial staking provider for ETH, managing infrastructure for over 500 institutional clients including asset managers and exchanges. Their Series B round brought in $50 million led by Senator Investment Group and Liberty City Ventures, with participation from Anchorage Digital and Galaxy Digital. This capital expands their proof-of-stake infrastructure across the Web 3 stack, following a smaller $2.5 million seed round. I view this funding as a clear signal that institutional demand is outpacing legacy capacity, forcing providers to scale their validator networks before yields compress further. The company predicts this segment will double overnight as Ethereum rails absorb more DeFi and NFT activity.

Infrastructure depth versus wallet reach

Figment’s technical edge lies in its developer-centric tooling, specifically the Hubble suite for analytics and the DataHub API supporting over 50 networks. They charge a staking fee of approximately 10%, structured as 0% on Consensus Layer rewards and 30% on Execution Layer rewards. This model appeals to teams needing granular data access and custom integrations for complex applications. In contrast, Consensys focuses on user-facing software like MetaMask, which has over 30 million monthly active users, and Linea, a zkEVM Layer 2 solution. Consensys raised $450 million in 2022 led by ParaFi Capital, valuing the company at roughly $7 billion. Their IPO preparation includes Morgan Stanley and Goldman Sachs as joint bookrunners, aiming for a listing earliest in 2026. While Figment sells infrastructure to institutions, Consensys sells access to consumers. I find Figment’s approach more resilient because it avoids direct exposure to consumer sentiment swings that often hit wallet companies harder during bear markets.

Kiln does not appear in the provided sources, so I cannot compare its metrics directly. However, the market dynamics favor providers who can offer audited infrastructure and slashing protection without custodial risk. Figment’s non-custodial model ensures customers hold withdrawal keys, while the service retains only signing key access for safety. This distinction matters when comparing against centralized exchanges like Binance, which stakes over 3.2 million ETH but operates under different regulatory constraints. You already know that staking yields have compressed to around 2.63% gross, making efficiency the primary driver for institutional choice.

Yield compression and operational reality

The staking landscape is defined by falling returns as more capital enters the system. Ethereum closed the third quarter with a record 43.8 million ETH staked, representing 35.9% of total supply. Gross yield slipped to 2.63% from 2.81% in the prior quarter, a decline of 0.18 percentage points. Validator counts fell to 881,000 from roughly 1.1 million due to consolidation after the Pectra upgrade, which allows validators to earn rewards on balances up to 2,048 ETH. This consolidation means fewer records do not indicate less security, but rather higher efficiency per operator. Bitwise noted that growth got cheaper by design, driven by issuance schedules rather than demand drops. Institutional players like BitMine increased their stake from 410,000 tokens to 5.07 million, driving much of the overall increase.

I believe Figment’s focus on multi-chain support gives it an advantage over single-asset specialists. Their ability to migrate between clients like Geth and Lighthouse quickly optimizes performance and safety. Consensys, meanwhile, benefits from the SEC dropping its lawsuit against MetaMask’s staking feature in February 2025, improving its regulatory standing. The question remains how much further yields can fall before staking growth stalls. With 35.9% of ETH locked, a growing share of supply is illiquid by design, unable to react quickly to price moves. This structural shift forces institutions to prioritize long-term network exposure over short-term income generation.