Nvidia, Stripe Lead $26 Billion Buying Spree Into Open-Weight AI
Three deals worth a combined $26 billion - including a reported $13 billion Nvidia bid for Hugging Face - have made open-weight AI infrastructure the hottest acquisition target in the Valley.
👁 115 reads
Nvidia is reportedly acquiring Hugging Face for $13 billion, a deal that - if confirmed - would cap a remarkable stretch in which open-weight AI infrastructure attracted more than $26 billion in acquisition capital within weeks. The reported bid follows Nvidia’s $6 billion agreement with Poolside, an open-weight model builder, under which most of Poolside’s employees will move to the chip giant. Two weeks earlier, Stripe paid more than $7 billion for OpenRouter, the leading provider of open-weight models to business customers.
- Nvidia is reportedly acquiring Hugging Face for $13 billion, following its $6 billion deal for open-weight model builder Poolside
- Stripe acquired OpenRouter, the top open-weight model provider to businesses, for more than $7 billion two weeks ago
- Only 6% of companies currently use open-weight models according to Ramp spending data, but cost pressures from frontier labs are pushing that figure up
The wave of deals reflects Nvidia’s need to reduce its dependence on hyperscalers and frontier labs at a moment when model builders such as OpenAI are designing their own inference chips - OpenAI’s Jalapeño was announced this week. By controlling the largest developer space for open models in the US, Nvidia gains direct access to a large user base it can direct toward its own chips and standards. Stripe, for its part, framed the OpenRouter purchase around compute economics. “Tokens are the central currency for companies building with AI, and it’s clear that the real-world economic potential will depend on making good use of scarce compute resources,” CEO Patrick Collison said in a statement.
Adoption of open-weight models remains narrow: just 6% of companies use them according to a Ramp survey of spending data, and only 2% of software engineers do so according to Jellyfish. Nik Albarran, AI product lead at Jellyfish, told TechCrunch that open models are currently used mainly for high-volume, repetitive workloads such as customer service, where tuning cuts inference costs. “If the prices continue to go up from the frontier labs, more and more companies will be forced to at least consider it,” Albarran said.
Google · Your sourcesQuantslant higher in your Google - one clickAdd to preferred sources →
More on this from QUANTSLANT
news·Judge Rules Pentagon’s Label Against Anthropic Was Illegal