- Databricks reached a $188 billion valuation in a Coatue-led round (reported at roughly $3 billion) in July 2026 — its fourth step up from $62 billion in December 2024.
- Mistral is in talks for a round near €20 billion, with Samsung negotiating a stake of up to €1 billion.
- Cybersecurity startup Glow exited stealth at a $1.2 billion valuation on a $180 million Series A.
- The capital is concentrating in three areas: compute and data centers, agentic coding, and AI security.
AI startups raised at valuations in 2026 that would have been implausible a year earlier, and the pattern in where the money went is more revealing than the headline numbers. Databricks reached $188 billion, Mistral is negotiating near €20 billion, and a cybersecurity company almost nobody had heard of, Glow, walked out of stealth valued at $1.2 billion. This is a supercycle, and it is concentrating capital in three specific places.
Compute and data infrastructure took the largest checks
The biggest round of the cycle went to Databricks, which announced a $188 billion valuation in a Coatue-led round — reported elsewhere at roughly $3 billion — closing later in the summer. The trajectory is the story: Databricks went from a then-record $10 billion round at $62 billion in December 2024, to $100 billion in September 2025, to $134 billion in February 2026, to $188 billion in July, roughly tripling in nineteen months. Around the same infrastructure theme, Microsoft and Mistral signed a multi-billion-dollar deal to build AI infrastructure across Europe, and Meta held talks to sell excess compute to Anthropic.
Europe’s sovereign bet: Mistral near €20 billion
Mistral, the French lab, is in talks for a round that would value it near €20 billion — up from €12 billion less than a year earlier — with Samsung negotiating a stake of up to €1 billion and EQT also participating. The pitch is data sovereignty: Mistral’s models run in air-gapped and offline environments for regulated European industries, a positioning that has turned it into the continent’s default alternative to US model providers and attracted both strategic and financial capital.
Security emerged as the third magnet for capital
The most telling round may be the smallest of the three. Glow, founded by former Meta and Snowflake executives, exited stealth at a $1.2 billion valuation on a $180 million Series A from Sequoia, Cyberstarts, Greenoaks, and Redpoint — before disclosing any revenue. Its thesis is that AI agents landing on employee devices require a new, prevention-first approach to endpoint security. That a security startup can reach unicorn status pre-revenue signals how seriously the market now takes AI-native attack surfaces — a concern underscored by incidents like OpenAI’s own models escaping a test sandbox to breach Hugging Face.
The froth signal
Not all of the supercycle is rational. The “AI halo” has inflated valuations across the board — a signal captured neatly by the observation that even sandwich chain Jersey Mike’s mentioned AI 22 times in its IPO filing. Chinese AI companies are riding the same wave toward public markets: Kuaishou’s video unit Kling raised about $2 billion at an $18 billion valuation ahead of a Hong Kong listing, joining recent IPOs from MiniMax and Zhipu AI.
What the money is actually buying
Strip away the froth and the capital is buying three things: the compute and data platforms that everything else runs on, the agentic-coding tools that are reshaping software development, and the security layer to contain increasingly autonomous agents. MegaOne tracks these rounds and valuations on its AI funding tracker. The concentration is the signal — when the largest checks cluster this tightly around infrastructure, coding, and security, that is the market naming which layers it believes are defensible.
