Analysis
More than 40% of seed and Series A capital deployed in 2026 has gone into rounds of $100 million or more, according to Crunchbase News, with roughly 62 such rounds closing so far this year worth a collective $15 billion.
The Biggest Prints
- [River AI](/pulse/company/river) -- $1.2B Series A: Silicon Valley platform for developers to train and serve custom models, a round Pulse has covered previously.
- Xpeng Robotics -- $900M Series A: China-based developer of AI-enabled humanoid robots.
- Apptronik -- $520M Series A: Austin-based humanoid robotics startup.
- Nexthop AI -- ~$500M Series A: AI infrastructure developer.
- MatX -- ~$500M Series A: semiconductor startup building AI accelerator chips.
- Mind Robotics -- ~$500M Series A: industrial robotics-focused startup.
“- Xpeng Robotics -- $900M Series A: China-based developer of AI-enabled humanoid robots.”
Why This Stage, Why Now
Jumbo Series A rounds concentrate in the categories that are most capital-intensive to build -- AI infrastructure, custom silicon and humanoid robotics all require significant upfront spending on compute, hardware and manufacturing before a company can prove out its product, which pushes what used to be several years of staged Series A-through-C fundraising into a single oversized check. Roughly half of this year's $100 million-plus Series A rounds went to US-based startups, with the rest concentrated in China and a handful of other markets building competing AI and robotics capacity.
The Numbers In Context
Four rounds of $500 million or more closed in Q1 2026 alone, a pace that would have been extraordinary at any point before the current AI investment cycle. Crunchbase's own tracking shows robotics and AI-model startups led new unicorn creation this year -- the jumbo Series A trend and the unicorn-minting trend are effectively the same phenomenon measured two different ways, since a $500 million-plus Series A at any reasonable multiple immediately creates a unicorn valuation.
The risk in this pattern is straightforward: Series A has traditionally been the stage where a company proves early product-market fit before committing to expensive scaling, and jumbo rounds compress that proof period, betting the underlying technology works before the same evidence a normal-sized Series A would require has been established.