The second week of August delivered the busiest stretch of AI dealmaking the Bay Area has seen since spring. Between August 8 and 14, three local companies alone pulled in more than $6.2 billion: Databricks closed a $5 billion round at a $190 billion valuation, Igor Babuschkin’s two-month-old River AI raised $1.1 billion across a combined seed and Series A, and CodeRabbit crossed into unicorn territory with a $143 million Series C. Per Crunchbase, the week’s ten biggest US rounds also reached into grid storage, defense drones, and datacenter networking - but the center of gravity never left AI, and it never left the Bay. AINsf collects the details.
Databricks: they wanted $1 billion, investors offered $15
The headline deal of the week was also the strangest negotiation. According to TechCrunch, Databricks set out to raise about $1 billion; investors wanted to put in $15 billion; the company settled on $5 billion at a $190 billion valuation. The round was led by Coatue, with Blackstone, MGX, and accounts advised by T. Rowe Price participating, and new backers including Sixth Street Growth, BOND, Clearlake Capital, Point72, Premji Invest, and TPG.
The numbers behind the demand: the San Francisco company says it crossed a $7 billion annualized revenue run rate in its second quarter, up more than 80% year over year, and has been cash-flow positive on an adjusted basis for each of the past twelve months. The raise comes roughly half a year after Databricks took in $5 billion at a $134 billion valuation - a 42% markup in about six months. CEO Ali Ghodsi told CNBC an IPO is still part of the plan, while downplaying any urgency on timing. For a company we last covered on stage at Moscone in June for its Data + AI Summit, the message is consistent: why price yourself in public markets when private ones keep doing this?
River AI: $1.1 billion for a company that is two months old
If the Databricks round says something about revenue, the week’s second-largest AI deal says something about faith. River AI, the Palo Alto startup founded this summer by xAI co-founder Igor Babuschkin, announced $1.1 billion across a Series Seed and Series A led by General Catalyst and AMP PBC, with strategic investment from NVIDIA and AMD Ventures and participation from Y Combinator and Temasek. TechCrunch notes the company is roughly two months old - making this one of the largest sums ever raised by a startup that young.
The pitch is an “open AI stack”: LoRA fine-tuning and reinforcement learning on frontier open-weight models, delivered through an API, so developers and enterprises can train, tune, deploy, and serve custom models without building an infrastructure team - and, crucially, keep the resulting weights. It is a direct bet against the closed-model API business, made with chipmaker money on the cap table.
CodeRabbit: the unicorn that reviews the robots’ homework
The most quietly telling deal of the week came from Walnut Creek. CodeRabbit, which sells AI code review, raised a $143 million Series C at a $1.5 billion valuation, co-led by Atomico and Smash Capital, with Datadog and BMW i Ventures among the new investors. The company says revenue grew more than 5x year over year, that it now runs over 2 million code reviews a week, and that more than 17,000 customers - including NVIDIA, BMW, Adyen, and Indeed - use it. Alongside the round it introduced Agentic Change Management, a governance layer for software written by people and agents, and pledged over $10 million to keep its review tools free for open-source maintainers for the next year.
The logic of the business is almost circular in the best way: the more code AI agents write, the more someone has to check it, and increasingly that someone is also an AI. A year in which coding agents went mainstream has produced, as its downstream effect, an East Bay unicorn whose product is reading their output.
The rest of the tab
The week’s megadeal list ran well beyond these three. Form Energy raised $750 million for grid-scale iron-air batteries - the energy bill of the AI buildout coming due. Neros Technologies took $250 million for defense drones, and San Jose’s Point2 Technology added $136 million for datacenter interconnects. Across the Atlantic, Stockholm’s Lovable - the vibe-coding startup - raised $400 million at a $13.3 billion valuation. The pattern across all of it is the one the Databricks negotiation made explicit: for the AI companies that are working, the constraint is no longer capital. It is how much of it they are willing to accept.