Key Points
PitchBook's Q2 2026 US VC data shows AI took 87.5% of all venture dollars in the quarter. Non-AI startups competed for 12.5% of available US venture capital in Q2 2026. Lovable raised $400M at a $13.3B valuation; River AI raised $1.1B in the same week. CodeRabbit closed $143M at a $1.5B valuation; Cognition AI is in talks at a $40B mark. The concentration is the most skewed AI-versus-non-AI split PitchBook has recorded.
US venture capital in Q2 2026 flowed almost entirely into artificial intelligence companies. Every other sector competed for the remainder. The numbers, when laid out plainly, are striking.
PitchBook released its Q2 2026 US VC Valuations report on Aug. 12.
87.5% of all US venture dollars went to AI. The remaining 12.5% was divided across every other technology sector.
The Numbers Behind The Concentration
The 87.5% figure covers all AI-related venture activity, from foundation model labs down to application-layer developer tools. It includes follow-on rounds, growth equity, and early-stage deals. PitchBook did not define a single breakout winner.
The data reflects a structural shift in how US institutional capital allocates to private technology companies.
A single week in August 2026 illustrates the data concretely. Lovable, the Swedish vibe-coding startup, raised $400 million at a $13.3 billion valuation. River AI, an open-weight infrastructure company, raised $1.1 billion across seed and Series A tranches just two months after leaving stealth.
CodeRabbit, an AI code review platform, closed a $143 million round at a $1.5 billion valuation, confirmed by Reuters. Cognition AI, the coding agent startup behind the Devin product, is in early discussions for a new round that could push its valuation past $40 billion, according to Bloomberg.
Four separate companies. Four rounds. All AI. All in a single five-day window.
Also Read: Grok 4.6 Scores 61 On Intelligence Index, Matching OpenAI's GPT-5.6 Sol
What Non-AI Companies Faced
The inverse of the AI concentration is a funding drought for companies that do not carry an AI narrative.
PitchBook's Q2 data shows that Series D and later-stage rounds have become particularly difficult for non-AI software companies. Investors are extending existing bets in AI rather than rotating into new sectors.
Biotech and climate tech, which absorbed significant capital in 2022 and 2023, saw reduced share. Consumer startups faced the sharpest contraction in dollar terms.
The dynamic is self-reinforcing. Large language model capabilities have improved fast enough that investors expect AI-adjacent tools to grow revenue faster than traditional software. Lovable's $500 million annualized run rate, reached in June, was cited as evidence. CodeRabbit did not disclose revenue in its announcement.
Non-AI founders have adapted by adding AI features to existing products, sometimes substantively and sometimes cosmetically. PitchBook analysts noted in their Q2 commentary that "AI" as a self-descriptor now appears in pitch decks at a far higher rate than actual AI-native architecture justifies.
Also Read: Claude Adds Invisible Watermarks To AI Text Under EU Transparency Rules, Users Push Back
What This Means For Crypto AI Crossover Projects
The AI capital concentration has a secondary effect on crypto-native AI projects. Decentralized compute and AI agent infrastructure startups compete for investor attention in a market where centralized AI labs are raising at valuations that dwarf most blockchain projects. A $1.1 billion River AI seed round is larger than the market capitalization of many established DeFi protocols.
That gap creates pressure on decentralized AI projects to demonstrate differentiated capabilities that centralized infrastructure cannot replicate. Privacy-preserving inference, censorship resistance, and permissionless access are the leading arguments. Whether those properties attract enough revenue to compete with centralized alternatives at scale is the open question heading into the second half of 2026.
Coinbase CEO Brian Armstrong posted on August 12 that AI agents are already onboarding as customers through Coinbase for Agents, citing organic growth without paid marketing.
The comment points to one area where crypto infrastructure may retain an edge over centralized AI platforms: programmable payments that AI agents can execute autonomously without human account setup.
The PitchBook data does not separate on-chain AI projects from centralized AI companies. If the category were subdivided, decentralized AI would represent a small fraction of the 87.5% figure.
The capital concentration, at least for now, sits firmly with closed-source, centralized labs and their application-layer offshoots.
Read Next: BIP-110 Failure Shows Why Bitcoin’s Original Launch Cannot Be Repeated
