Inferact

AI infrastructure startup commercializing vLLM, the open-source LLM inference engine created at UC Berkeley's Sky Computing Lab. vLLM supports 500+ model architectures across 200+ accelerator types and runs on 400,000+ GPUs concurrently worldwide, with production users including Meta, Google, Character.ai, and Amazon. Inferact is building what it calls a "universal inference layer" that works with existing inference providers rather than competing against them. Emerged from stealth in January 2026 with a $150M seed round at an $800M valuation, co-led by Andreessen Horowitz and Lightspeed, with participation from Sequoia, Altimeter, Redpoint, and ZhenFund. Founded by vLLM maintainers Simon Mo (CEO), Woosuk Kwon, Kaichao You, and Roger Wang.