Sources: AI inference startup Baseten raised $300M led by IVP and CapitalG at a $5B valuation, more than doubling its previous valuation; Nvidia invested $150M
The move follows other investments from the chip giant to improve the delivery of artificial-intelligence services to customers
Context & Ripple Effects
Baseten’s reported $300M round marks a sharp step-up from its earlier $40M Series B for deploying open-source and customized models and the $75M round that valued it at $825M. IVP participated across that progression, while CapitalG joins as a reported lead in this round.
The financing puts a specialist inference provider at the intersection of venture funding and Nvidia’s effort to support delivery layers for AI services, not only the chips beneath them.
First-order effects
- Baseten gains $300M of reported financing and a $5B valuation, giving it more capital to build and sell its inference offering; IVP and CapitalG become the round’s reported lead backers.
- Nvidia’s reported $150M investment gives Baseten a strategically aligned investor as it competes to serve customers deploying AI models.
Second-order effects
- Other inference providers face a higher financing and credibility bar: a well-capitalized Baseten can invest more aggressively in product delivery and customer acquisition.
- Nvidia’s participation reinforces the commercial importance of the layer that turns AI compute into customer-facing services, potentially concentrating attention and capital on providers with close hardware-ecosystem ties.
Third-order effects
- If such investments continue, value creation in AI infrastructure may shift beyond chip supply toward inference platforms that control how models are deployed, optimized, and delivered to customers.
- The pattern could deepen AI infrastructure financialization: large strategic and growth investors may increasingly determine which compute-adjacent platforms can fund the scale needed to compete.
The trend: This is one data point in the race to capture recurring value from AI inference services rather than from underlying compute alone.