Presented by
Greenland Bitcoin Bump - Trump stated he is seeking "immediate negotiations" for Greenland acquisition and confirmed he would not use force, noting "This will not be a threat to NATO."
DePin Hype Fades - Venture investors are increasingly divided over the Decentralized Physical Infrastructure Networks (DePIN) sector as scrutiny shifts from its initial $3.5 trillion market projection to fundamentals like capital costs and cash flow.
Saylor Buys BTC - MicroStrategy CEO Michael Saylor announced the acquisition of an additional $2.13 billion worth of Bitcoin.
New here? Ask Copilot up to 5 free queries per day with Messari Basic.
Messari's protocol reports give you a deep dive on the foundation and state of top crypto protocols, including key metrics and notable events. See the complete list of protocol reports here and get a preview of our latest report below.
Network activity reached new heights in Q4, with average daily transactions increasing 13.7% QoQ to 10.2 million and average daily active addresses growing 12.3% to 2.8 million.
In Q4 2025, TRON experienced a contraction in revenue, falling below the billion-dollar mark achieved in the previous quarter. Total revenue in USD decreased by 38% QoQ, dropping from $1.1 billion to $655.6 million.
The stablecoin ecosystem continued its steady expansion, with TRON’s stablecoin market cap increasing 7.0% QoQ to $81.8 billion. USDT remains the dominant asset with a $80.9 billion market cap, representing over 99% of the network's stablecoin supply.
Ecosystem momentum was driven by major institutional and cross-chain integrations, including native staking on Ledger Live, TRX launching on the Base network via LayerZero, and USDT on TRON being cleared by the Abu Dhabi Global Market (ADGM) for regulated use.
SunPerp marked a successful first full quarter, achieving over $25 billion in total trading volume since its launch, utilizing a hybrid model to offer zero-gas-fee trading and high-speed execution.
Wormhole connects over 40 blockchains and has processed more than $70 billion in cumulative cross-chain volume and one billion cross-chain messages, making it one of the most widely used interoperability protocols in crypto.
Wormhole’s NTT (Native Token Transfers) standard aims to replace fragmented wrapped assets with canonical multichain representations, consolidating liquidity, preserving issuer control, and simplifying risk management.
NTT is already in production across multiple deployments, including Ripple’s RLUSD rollout, Dogecoin’s canonical DOGE implementation, M0’s multichain M token, and Monad’s native bridge.
Wormhole Portal has evolved from a standalone bridge interface into a multichain gateway, integrating cross-chain swaps via Mayan.
Institutional adoption spans tokenization and real-world asset platforms (e.g., BlackRock, Ripple, Apollo, Hamilton Lane, Mercado Bitcoin, Centrifuge, Securitize), which use Wormhole for cross-chain distribution of tokenized funds, tokenized private credit, and stablecoins.
Bitget’s Universal Exchange (UEX) model combines centralized efficiency, decentralized accessibility, and traditional financial product offerings into a unified trading experience for retail and institutional participants. The UEX model comprises the core centralized exchange, Bitget Onchain, Bitget Wallet, Morph, and GetAgent.
Tokenized stocks are emerging as a high-growth vertical in the Bitget ecosystem. Since launching in July, a total of $17.1 billion in tokenized stock futures volume has been generated.
Bitget Onchain is the primary mechanism for integrating multichain access into the platform. Launched in April 2025, it enables users to trade onchain assets using CEX balances across multiple networks, generating over $2.4 billion in cumulative volume.
GetAgent is the platform’s AI assistant, designed to simplify the user experience as product coverage expands. Beyond market insights and guided execution, GetAgent has been used as a distribution channel for engagement campaigns.
From Our Sponsor
As AI shifts from offline generation to interactive systems (avatars, agents, live analytics, world-model gaming), the bottleneck is no longer model quality. It's whether systems can run continuous, frame-by-frame inference under tight latency constraints. That workload shape is fundamentally different from the batch jobs most GPU infrastructure was designed around, and it’s creating a new category boundary: real-time AI video as a first-class infrastructure problem.
Livepeer is a clean signal of that boundary forming. By the end of 2025, AI inference accounted for approximately 70% of protocol demand-side fees on the network, reflecting a sustained shift away from traditional transcoding toward real-time, AI-driven video workloads.
Rather than expanding toward generic inference, Livepeer’s Cascade vision reframes the network around real-time AI video pipelines, composable workflows that operate on live streams rather than static files. This is not a branding tweak; it is a constraint-driven repositioning toward a niche where decentralized execution is most defensible: workloads that are continuous, media-native, and coordination-heavy.
From “Protocol” to “Product Surface”
This is where Livepeer’s shift becomes visible in practice.
Daydream as the wedge: launched as a real-time creative experience, Daydream makes the constraint visible (latency, throughput, and reliability) and turns it into a developer-facing workflow surface.
Productization as the strategy: Livepeer has been explicit that the focus is productizing real-time AI video workflows rather than competing as a broad, general-purpose compute platform.
Roadmap as coordination: the Livepeer Foundation published a phased roadmap (“Now / Next / Beyond”) oriented around performance, GTM, and enterprise readiness, i.e., the prerequisites for live pipelines to become durable demand.
Why this matters for DePIN x AI
DePIN’s first wave showed that token incentives could rapidly bootstrap infrastructure supply, which initially attracted investor interest. As these networks matured, the harder problem became clear: converting capacity into sustained, paid utilization.
That dynamic is beginning to shift as certain AI workloads prove structurally compatible with decentralized infrastructure. Livepeer’s trajectory suggests that demand for real-time AI video is not hypothetical, but the remaining challenge is operational execution rather than demand discovery.
Real-time AI video is one of the first AI segments that naturally aligns with decentralized networks. The workload is continuous, latency-sensitive, and operationally constrained, favoring elastic capacity, distributed operators, and open coordination over static provisioning models.
If this category scales, success will not be defined by access to the lowest-cost GPU. It will be defined by the ability to deliver production-grade real-time pipelines, with measurable latency, predictable routing, repeatable workflows, and developer primitives designed for reliability rather than experimentation.
