Infrastructure-Heavy AI and Asset-Backed Climate Finance Reshape the Weekly Venture Leaderboard

Executive Summary

The week’s defining story wasn’t merely that capital flowed—it's where it concentrated: into AI infrastructure (inference, networking, photonics) and asset-backed climate/platform financing that looks increasingly like structured credit markets borrowing venture’s narrative layer. The clearest example is a clean-energy installer platform pairing a relatively modest equity round with €1.02B of debt plus a €300M EIF guarantee, effectively turning “climate software” into a balance-sheet optimization business. In parallel, AI’s gravity pulled late-stage money into picks-and-shovels (inference platforms, AI networking) while also minting one of the largest seed rounds on record for a new AI research lab.

Marker size: 1–5 deals 6–15 16–30 30+

Across 150 deals totaling $8.1B, the mix signals a market that is risk-on in dollars but selective in underwriting: very large checks are being written where investors can point to either (a) infrastructure bottlenecks (compute, inference, networking), (b) regulated/mission-critical outcomes (healthcare AI, OT security), or (c) cash-flow visibility and collateral (fiber, securitized music rights, debt-backed energy financing). By count, early-stage activity remains broad (45 seed deals; 33 Series A deals), but by dollars the center of mass is still Series C+ ($3.3B) and a small number of mega rounds.

Three themes stood out: 1. AI funding is bifurcating into foundation-lab ambition at the seed stage and production/infrastructure at growth. The seed mega-round for an AI research startup and the growth rounds for inference and AI networking highlight investors’ desire to own both the “brains” and the “pipes.” 2. Climate and industrial policy are meeting private credit. Large debt facilities, guarantees, and manufacturing capacity expansions (rare earth magnets; geothermal buildout) reflect a financing playbook more akin to infrastructure funds than traditional VC. 3. Healthcare is rewarding adoption and workflow ownership, not just models. Medical search with physician distribution, obesity/chronic care with outcomes evidence and profitability, and a brick-and-click pediatrics roll-up point to a 2026 thesis: own the clinical front door, then layer AI.

What this signals: 2026 is shaping into a year where venture returns will be increasingly determined by capital structure literacy (who can combine equity with debt, securitization, guarantees, and project finance) and by distribution advantage (who has users, clinicians, or enterprise customers already embedded). The “valuation reset” narrative is fading in these pockets—several rounds doubled valuations or posted sharp step-ups—yet underwriting remains anchored in tangible scaling vectors: deliveries completed, clinics opened, passings built, consultations served, or revenue projections tied to financed assets.

---

Mega Deals & Headline Rounds (>$50M)

Cloover — €1.04B Series A (Germany)

Investors: Lowercarbon Capital, QED Investors, Bosch Ventures, MMC Ventures, BNVT Capital, Earthshot Ventures, Centrotec
What happened: A blended financing stack: €18.8M Series A equity led by MMC Ventures and QED plus a €1.02B debt facility from a major European bank and a €300M guarantee from the European Investment Fund.

Cloover’s round is best read as a climate-finance platform deal rather than a traditional software Series A. The company serves installers across Germany, Switzerland, Sweden, the Netherlands, and Austria—markets where distributed energy (solar, heat pumps, batteries) is constrained less by demand than by installation throughput, working capital, and consumer financing. By pairing installer workflow software with financing rails, Cloover becomes a “merchant of record” for project economics, pulling forward adoption by making installs financeable and standardized.

The investor set is unusually coherent: Lowercarbon pushes the climate deployment thesis, QED brings fintech underwriting DNA, Bosch Ventures signals strategic pull in hardware/installation ecosystems, and MMC fits European growth/AI-enabled SaaS. The market implication is that 2026 climate winners may increasingly be those who can industrialize project origination—then finance it at scale. Cloover projecting €426.7M revenue in 2026 and €850M in 2027, while claiming profitability and 8x revenue growth in 2025, suggests that “climate software” is entering the realm of high-velocity credit businesses where cost of capital, defaults, and installer retention matter as much as product.

Zipline — $600M (United States)

Investors: Baillie Gifford, Fidelity Management & Research, Valor Equity Partners, Tiger Global
Key points: $7.6B valuation, expansion to Houston and Phoenix and at least four U.S. states in 2026; 2M+ commercial deliveries; 15% WoW U.S. delivery growth for seven months; median delivery time 3 minutes.

Zipline’s financing looks like public-market crossover money re-embracing a scale story with operational proof. Deliveries at this level—paired with aggressive growth rates—shift the narrative from “promising robotics” to “repeatable logistics network.” The median 3-minute delivery time is not a vanity metric; it’s a proxy for service-level advantage that can unlock enterprise and healthcare workflows (urgent meds, diagnostics) and potentially consumer convenience.

The investor roster matters: Baillie Gifford and Fidelity are long-duration growth investors, while Valor is known for operational scaling bets, and Tiger continues to show selective appetite for category leaders. The market implication is that autonomous logistics is moving from R&D to network buildout, and the winners will be those who secure regulatory pathways, landing sites, and enterprise contracts fast enough to justify capex. Zipline’s round is a signal that late-stage capital will still pay up for real-world deployment curves.

Humans& — $480M Seed (United States)

Investors: NVIDIA, GV, Emerson Collective, SV Angel, Jeff Bezos, Georges Harik
Key points: Founded ~3 months ago; alumni from Anthropic, xAI, Google, Stanford; $4.48B valuation; heavy compute spend planned.

This is a “seed round” in label only; functionally it is a pre-formation capitalization for a frontier AI lab. The valuation implies investors are paying for (1) team pedigree, (2) a differentiated research thesis (“human-centric AI for collaboration and connection”), and (3) presumed access to compute and distribution networks. The most important tell is NVIDIA’s participation—both as capital and as a strategic signal that compute allocation and model training infrastructure remain a gating factor.

The round also reflects a market psychological shift: after a period of consolidation anxiety, investors are again underwriting new labs—but only those with credible founders and clear compute plans. The implication for incumbents is competitive: deep-pocketed new entrants will bid for the same scarce inputs (research talent, data partnerships, and GPU time). For founders outside frontier labs, this also raises the bar: application-layer AI must increasingly show workflow capture and monetization, not just model novelty.

Baseten — $300M Series C+ (Canada)

Investors: NVIDIA, CapitalG, IVP
Key points: Co-led by IVP and CapitalG; NVIDIA invested $150M; valuation doubled to $5B.

Baseten sits at the heart of the “inference economy”: deploying, scaling, and serving models reliably in production. The valuation doubling indicates that the market is rewarding revenue traction in production AI, where reliability, cost-per-token optimization, latency, and governance matter more than model benchmarks.

Strategically, NVIDIA writing a $150M check is both a vote of confidence and a distribution accelerant: inference platforms that align with NVIDIA’s stack can become default pathways for enterprises. CapitalG’s involvement fits an enterprise software infrastructure thesis, while IVP tends to fund category leaders in scaling mode. The competitive landscape includes hyperscaler-managed services and a proliferating set of inference startups; Baseten’s path to defensibility likely hinges on developer experience + performance + multi-cloud flexibility.

Mews Systems B.V. — $300M Series C+ (Netherlands)

Investors: Battery Ventures, Atomico, Kinnevik, Tiger Global, HarbourVest Partners, EQT Growth
Key points: Series D led by EQT Growth; values at $2.5B post; AI-native features, automation agents, fintech expansion (Mews Payments), North America + Europe growth.

Mews is a reminder that vertical SaaS can still command large growth rounds when it adds a credible fintech layer. Hospitality management systems are sticky, but historically low-ARPU; payments can change the unit economics by monetizing GMV. This financing suggests investors believe Mews can become the operating system + financial layer for hotels, benefiting from post-pandemic digitization and labor automation.

Atomico and EQT Growth imply a European champion-building agenda, while Tiger’s presence signals conviction in global scaling. The AI angle—agent-driven automation—matters because hotel operations have high variability (late check-ins, room assignment, maintenance). The implication: 2026 vertical SaaS winners will be those that ship automation with measurable labor savings and attach fintech to capture incremental margin.

Corxel Pharmaceuticals — $287M Series C+ (United States / China-linked)

Investors: RA Capital, Adage, HBM Healthcare, Invus, RTW, SilverArc, SR One, TCG Crossover, Hengdian Group Capital, others
Key points: Series D1 to advance CX11 through U.S. Phase 2; global Phase 2 for T2D; Phase 3 prep; HQ Shanghai with U.S. ops.

Corxel’s round shows that crossover biotech capital is back—selectively—around cardiometabolic therapies, a category with enormous TAM but high trial and commercialization costs. The syndicate is telling: RA, RTW, and Adage are experienced in late-stage biotech risk, while TCG Crossover signals public-market optionality.

The cross-border footprint matters in 2026: China-linked biotech assets can still attract U.S.-anchored capital when clinical strategy is U.S.-centric and IP/licensing is structured carefully (here, ex-China rights were licensed from Vincentage). Investors are effectively underwriting data readouts + regulatory execution rather than platform narratives.

OpenEvidence — $250M Series C+ (United States)

Investors: DST Global, Thrive Capital
Key points: Widely used AI platform by U.S. doctors; 18M clinical consultations in Dec 2025; free for physicians (NPI verified), ad-supported; valuation doubled to $12B; nearly $700M raised in last 12 months.

OpenEvidence is an archetype of “distribution-first healthcare AI.” The usage metric—18M consultations in a single month—suggests the product is embedded in clinical workflow at a scale competitors will struggle to match. The ad-supported model is unusual for clinical software but potentially powerful: it enables rapid adoption while building a defensible audience for pharma/health advertisers.

DST and Thrive are signaling that healthcare AI can achieve consumer-like scale inside a professional market if it nails trust, verification, and speed. The key risk is regulatory and reputational: medical information surfaces must be extremely reliable. But the market implication is clear: in 2026, the most valuable healthcare AI companies will be those that own a daily clinician touchpoint.

Noveon Magnetics — $215M Series C+ (United States)

Investor: One Investment Management
Key points: $200M led; expands domestic manufacturing beyond 2,000 tons/year; reshoring critical supply chains; secondary sales; two new board seats.

This is venture-adjacent industrial scaling driven by geopolitics and supply chain security. Rare earth magnets are critical for EVs, defense, and industrial motors—and the U.S. has prioritized domestic capacity. OneIM’s board control indicates a private equity-style posture, emphasizing governance and scale execution.

The implication: “hard tech” in 2026 is increasingly financed by investors comfortable with manufacturing ramp risk, and the equity story is intertwined with policy tailwinds and long-term offtake agreements.

Pennylane — $200M Series C+ (France)

Investors: TCV (lead), DST Global, CapitalG, Sequoia Capital, Blackstone Growth
Key points: Series E; valuation ~$4.25B; expansion in Europe (Germany), AI features, regulatory compliance (France e-invoicing); 6,000 accounting firms and 500,000 SMEs served.

Pennylane sits on a regulatory inflection: France’s e-invoicing mandate is a distribution catalyst, forcing digitization and creating a winner-take-most dynamic for platforms that can onboard accountants and SMEs. The investor stack—TCV + Blackstone Growth + Sequoia/DST/CapitalG—reads as “European ERP-in-the-making.”

Strategically, Pennylane is competing against legacy accounting suites and newer SMB finance tools. Its defensibility is in workflow integration with accounting firms (a channel moat) plus compliance functionality. The market implication: 2026 fintech growth rounds will increasingly cluster around regulatory-driven adoption and accountant-led distribution.

Upscale AI — $200M Series A (United States)

Investors: Tiger Global, Premji Invest, Xora Innovation (lead group), Intel Capital, Qualcomm Ventures, Maverick Silicon, Prosperity7, StepStone
Key points: Oversubscribed; total funding >$300M; unicorn within months of $100M seed; building AI networking solutions; active in UEC, OCP, SONiC; shipping later this year.

Upscale AI is effectively a bet that networking becomes the next AI bottleneck after GPUs. As clusters scale, interconnect and fabric efficiency can make or break training/inference economics. The consortium involvement (Ultra Ethernet, OCP, SONiC) suggests a strategy of embedding into emerging standards rather than fighting them.

Tiger’s participation here (and in Zipline and Mews) suggests thesis evolution: selective big bets on infrastructure and category leaders rather than broad spray. Intel and Qualcomm’s presence implies strategic alignment with silicon roadmaps. The implication: we’re entering a phase where “AI infrastructure” is not only compute—it's connectivity, orchestration, and hardware-software co-design.

Duetti — $200M Series C+ (United States)

Investor: The Raine Group
Key points: $50M equity + $125M securitization + $25M credit facility increase; accelerates music catalog acquisitions; 1,100+ artists; 80+ deals/month.

Duetti’s capital stack is the story: securitization turns streaming royalties into financeable cash flows. This is less a software company than a financialized rights platform with tech-enabled underwriting and sourcing. Raine’s media-finance expertise is a strong fit.

The implication is broader than music: venture is increasingly comfortable with structured finance attached to tech platforms. The risk is macro sensitivity—streaming payouts, interest rates, and catalog valuations can fluctuate.

Oviva — $200M Series C+ (Switzerland)

Investors: Planet First Partners, Kinnevik (led with $100M), Temasek, Sofina, Norrsken, Lunate, EGSB, A.P. Moller Holding
Key points: Expansion into hypertension/T2D; 1M patients; 90+ peer-reviewed studies; cash-flow profitable in 2025; tripled new patient intake in two years.

Oviva represents the “outcomes + evidence” path to durability in digital health. Profitability plus peer-reviewed validation is rare in the category and gives the company leverage with payors and health systems. Kinnevik doubling down underscores a European scale-up pattern: back the winners to become multi-condition platforms.

The market implication: GLP-1 era obesity care is catalyzing adjacent chronic-care platforms, but investors will favor those that can demonstrate clinically validated engagement and reimbursement pathways.

Surf Internet — $175M Series C+ (United States)

Investors: Bain Capital, Macquarie Capital, Future Standard
Key points: Completes prior $175M equity commitment; fiber buildout; targeting 330,000 passings in 2026.

Fiber is infrastructure financing disguised as venture. The underwriting is about build velocity, take rates, and unit economics per passing. Bain/Macquarie signal a capital-intensive play where scale and financing sophistication win.

Implication: as AI drives bandwidth demand and remote work persists, regional fiber operators can still raise large rounds—particularly when there’s a credible roadmap to passings and cash-flow.

Preply — $150M Series C+ (United States)

Investors: WestCap, Horizon Capital, EBRD
Key points: Series D; $1.2B valuation; AI capabilities; global growth; 100k+ tutors/learners across 180 countries; recently EBITDA positive.

Preply’s round shows the edtech rebound is selective: marketplace leaders with global liquidity and improving profitability can raise growth capital. AI features likely focus on tutoring augmentation, lesson planning, assessment, and retention.

Implication: edtech investors are shifting from “user growth” to unit economics + AI-enabled margin expansion.

Claroty — $150M Series C+ (United States)

Investor: Golub Capital (Golub Growth)
Key points: Valuation > $3B (80% increase since Mar 2024); supports expansion and M&A; 24 of Fortune 100 as customers; AI-powered CPS Library with Schneider Electric and Rockwell.

OT security is benefiting from rising cyber-physical threats and regulatory scrutiny. Golub’s role suggests a growth-credit/structured equity flavor: capital for scaling plus potential acquisitions in a consolidating market.

Implication: cybersecurity funding is gravitating to mission-critical domains (OT/CPS) where budgets are resilient and switching costs are high.

Inferact — $150M Seed (Canada)

Investors: Altimeter, Redpoint, Lightspeed, a16z, Databricks Ventures, ZhenFund, Sequoia, UC Berkeley Chancellor’s Fund
Key points: $800M valuation; evolved from open-source vLLM (UC Berkeley); led by Simon Mo and Ion Stoica; commercial serverless + optimization.

Inferact is a clean example of open-source “research-to-company” acceleration. vLLM’s adoption makes this a distribution-first seed: investors are buying into an existing developer footprint and a team with deep credibility. The valuation reflects the market belief that inference optimization is a massive spend category.

Implication: expect more open-source maintainers to raise very large seeds when they control critical infrastructure primitives.

Zanskar — $115M Series C+ (United States)

Investors: Spring Lane Capital (lead), Lowercarbon, Munich Re Ventures, USV, Obvious, and many others
Key points: AI discovery platform + exploration drilling + plant construction; total equity now $180M; aims for power before 2030.

Geothermal is having a moment because it’s dispatchable clean baseload—valuable in grids dominated by intermittent renewables. Zanskar’s blend of AI exploration and physical buildout is ambitious; the syndicate spans climate VCs and energy/insurance capital (Munich Re), suggesting underwriting of both technology and project risk.

Implication: the next climate scale-ups will be those that can move from “software-enabled discovery” to owned or contracted generation assets.

Zarminali Pediatrics — $110M Series A (United States)

Investors: Healthier Capital (lead), General Catalyst, K2 HealthVentures
Key points: 28 clinics across 8 states; 15 new clinics in 2026; adding urgent and specialty care; Amir Dan Rubin joins board; total funding $150M.

This is a care-delivery roll-up with a software layer—an increasingly popular model as payors seek lower-cost settings and better patient experience. The speed of clinic rollout indicates operational sophistication; the board addition suggests a focus on scaling governance and payer strategy.

Implication: healthcare services are back in favor when they use capital to buy market presence and patient flow, then apply tech to improve margins.

Neurophos — $110M Series A (United States)

Investors: Alumni Ventures, Gaingels, M12, Bosch Ventures, Gates Frontier, Silicon Catalyst, Aramco Ventures, DNX, Morgan Creek, and others
Key points: Photonic AI chips; claims 10,000x miniaturization of optical modulators; building datacenter-ready OPU modules + software; Austin HQ expansion.

Photonics is one of the more credible paths to energy-efficient compute, but it’s notoriously hard to productize. The investor mix combines strategic and deep-tech capital; Silicon Catalyst and M12 imply ecosystem and go-to-market support.

Implication: as power and heat become binding constraints in datacenters, investors are taking more shots on non-CMOS compute adjacencies, but will demand credible packaging and software stack plans.

DealHub — $100M Series C+ (United States)

Investor: Riverwood Capital
Key points: Growth round to accelerate global expansion; “Agentic Revenue Hub.”

DealHub is a classic growth equity play in revenue operations: CPQ, billing, and contract lifecycle are systems of record where switching costs can be high. The “agentic” positioning suggests using AI to automate quoting, renewals, and approvals.

Implication: investors believe the next RevOps wave is AI-native workflows layered onto existing revenue infrastructure, creating upsell and cross-sell expansion.

---