The participant list on OpenAI’s $122 billion February 2026 round is a compressed map of the AI infrastructure economy. Amazon, Nvidia, and SoftBank wrote checks — three organizations whose strategic interests around OpenAI’s success are intertwined but distinct. Together with xAI’s $20 billion January close, those two deals pushed Q1 2026 generative AI venture capital to $145 billion, an all-time record according to S&P Global Market Intelligence.
That record is real, but the market beneath those two transactions tells a different story. Approximately $3 billion reached the rest of the generative AI startup landscape across the quarter. The median seed-stage AI valuation in March 2026 was 18% below where it sat in March 2025. Concentration and compression are happening simultaneously.
Nvidia’s Rationale
Nvidia supplies the GPUs that power OpenAI’s training and inference infrastructure — the H100 and forthcoming B100 series chips that have become the currency of foundation model development. OpenAI is among the largest single purchasers of Nvidia compute capacity in the world. An equity stake in OpenAI aligns Nvidia’s financial returns with the continued growth of its most important customer category. It also gives Nvidia visibility into OpenAI’s long-term compute roadmap, which is relevant to Nvidia’s own product planning and inventory commitments.
The investment is consistent with Nvidia’s broader move beyond pure chip sales. CEO Jensen Huang has positioned the company as a full-stack AI infrastructure provider, with software, networking, and now direct financial exposure to the model layer it powers.
The Amazon Dual Bet
Amazon’s position in the OpenAI round is the most strategically complex element of the Q1 capital activity. AWS has committed up to $4 billion to Anthropic, built custom silicon for Anthropic’s model training, and made Anthropic models a cornerstone of its enterprise AI offerings. Joining the OpenAI round creates an apparent conflict: AWS is now the primary infrastructure partner and financial backer of two companies whose models compete directly for the same enterprise customer budgets.
The logic, from Amazon’s perspective, is that enterprise customers will make their own model-provider choices regardless of what AWS prefers. If AWS is not financially and commercially aligned with both major providers, it risks losing the infrastructure revenue — and the data center economics — that follows to Azure or Google Cloud. Backing OpenAI is a defensive posture as much as a growth bet.
The Vertical AI Opportunity Below the Megadeals
The companies that make Q1’s record consequential over the long term are not OpenAI and xAI — they already exist. The consequential players are the applied AI companies that raised portions of the quarter’s remaining $3 billion and are now executing against enterprise sales plans in healthcare, legal, and financial services.
Series A and B rounds in those categories have continued at the $50 million to $200 million range, with economics that look like enterprise SaaS at accelerated growth rates. Multi-year customer contracts, high net revenue retention, and proprietary training data create the kind of defensibility that pure model builders cannot claim. These companies build on top of foundation models rather than competing with them — a position that insulates them from the existential competition OpenAI and xAI are waging against each other.
The talent question remains the central execution risk. Senior machine learning engineers are receiving offers from OpenAI and xAI that are difficult for a Series B company to match on equity alone. Founders who solve that recruiting equation and deliver on their revenue commitments over the next 12 months will produce outcomes that stand alongside the megadeals in the full history of this investment cycle — just measured differently, and paid back sooner.
Source: Generative AI Pulled In a Record $145 Billion in Q1 Venture Capital