bytevyte
bytevyte
Language
ai-beats

AI Funding Tracker: Two $3B Rounds and Nvidia's $2.5B Stake Set the Pace

This period's AI funding tracker covers twelve items, and the largest disclosed equity rounds land at $3 billion each. Crusoe's $3 billion raise set a $30 billion valuation, while Moonshot AI's Hong Kong filing targets roughly $3 billion at a reported $50 billion valuation. Close behind, Nvidia is negotiating a $2.5 billion stake in Thinking Machines Lab that would value the startup near $40 billion, with Accel in line to lead.

DateDealAmountRoundValuation
2026-09-04Nvidia circles a $2.5 billion stake as Thinking Machines Lab funding talks target a $40 billion valuation$2.5B$40B
2026-09-04Moonshot AI Hong Kong IPO: a $3 billion raise tests China AI's global capital route$3BIpo$50B
2026-09-06A $45B Anthropic Deal Pushed Nscale's Contracted Revenue Backlog to $103B$45BIpo
2026-09-07The $15B Anthropic pre-IPO credit facility is a fee disguised as a loan$15BIpo
2026-09-07Fluidstack $1.5B funding round, led by Jane Street, lifts neocloud valuation past $18B$1.5B$18B
2026-09-07Agility Robotics SPAC valuation: $2.5B on $1.8M in sales and a $140M loss$1.8M$2.5B
2026-09-08Crusoe $30 billion valuation puts contracted GPU revenue to the test$30B$30B
2026-09-08Uber's $100M stake backs reported Atoms robotaxi plans$100M
2026-09-09Google's Finland AI investment: €13B turns Nordic power into Europe's largest AI program$13B
2026-09-09Anthropic safety warning sharpens as pretraining researcher quits before October IPOIpo
2026-09-10Moonshot AI IPO Plan Puts a $50 Billion Price Tag on Chinese Frontier AI$50BIpo$50B
2026-09-11Altera IPO Filing Sets Up a $2 Billion Test of the Chip Listings Window$2BIpo

Compute capacity stayed the most expensive asset on the board. Nscale's contracted revenue backlog reached $103 billion after a $45 billion Anthropic deal, ahead of a possible September listing. Fluidstack drew a $1.5 billion round led by Jane Street that pushed its valuation past $18 billion, and Crusoe's $30 billion mark rested on a reported $13 billion Jane Street GPU contract. Across the AI funding tracker's compute names, pricing follows signed contracts rather than installed hardware, which leaves that backlog exposed to a single counterparty at Nscale and to GPU demand at Crusoe.

Compute Deals Dominate the AI Funding Tracker

Anthropic's own financing produced the period's largest single facility. Its pre-IPO credit line expanded to $15 billion with Morgan Stanley leading 14 banks, sized against a listing that could reach $2 trillion. For those lenders, the commitment buys positioning in a listing that would rank among the largest on record. Two safety data points landed in the same window: pretraining researcher Jacob Coxon resigned, and alignment lead Evan Hubinger placed extinction risk above 10%, both before an October IPO.

Listings Pipeline Widens

Three listings entered the pipeline. Moonshot AI's confidential Hong Kong filing pairs with a Shanghai listing, and the Kimi developer is courting Microsoft, Amazon and Google as backers. Altera filed confidentially for an IPO that could raise more than $2 billion, testing whether semiconductor listings have reopened for FPGA assets. Agility Robotics carries the period's widest gap between valuation and revenue: a $2.5 billion SPAC valuation against $1.8 million in 2025 net sales and a $140 million operating loss. That multiple, not the compute contracts, is the one public buyers will have to justify first.

Two items sit outside the round-and-valuation pattern. Uber's $100 million stake in Kalanick's Atoms adds weight to reported robotaxi plans, which the company denies despite its Pronto acquisition. Google committed €13 billion to four Finnish sites, backed by a 22-year nuclear agreement, wind power and a 94 MW battery.

Next period turns on execution rather than announcement. Nscale's September listing and Anthropic's October window would both show how much of the AI funding tracker's contracted backlog converts into public-market revenue.

✔Human Verified


Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.