Together AI raised $800 million in Series C funding on July 1, 2026, leaping to an $8.3 billion valuation. The neocloud infrastructure company rents out Nvidia GPU clusters and other specialized compute resources to enterprises and startups building AI systems.

The round was led by Saudi Arabia’s Aramco Ventures and included Vista Equity Partners, General Catalyst, Emergence Capital, Nvidia, and others. It marks a turning point: startups are no longer just collecting capital. They’re solving real infrastructure problems that limit AI deployment at scale.

Why Compute Layers Matter Now

A year ago, the question was whether AI would work. Now it’s how to run it cheaply at scale. Together’s bet is that enterprises don’t want to buy their own GPUs. They want to rent compute on demand, the way they rent cloud servers. Aramco’s participation signals that even oil-backed capital sees AI infrastructure as essential.

Nvidia benefits directly from the deal. Together buys Nvidia chips to lease back to customers. More Together customers means more Nvidia revenue. The investment pays dividends both ways.

The Crowded Space Gets Crowded

Together is not alone. Lambda Labs, Crusoe Energy, and others compete in the same lane. But Together’s focus on decentralized, consumer-friendly GPU access gives it a different angle than traditional cloud providers like AWS or Azure.

The $800 million valuation is large for an infrastructure play, but small in the context of where capital is flowing. Anthropic closed at $965 billion. OpenAI hit $852 billion. Together’s 51x smaller, but it operates in a less saturated market. If AI deployment becomes standard across industries, the demand for cheap compute is unlimited.

The timing matters: AI models keep growing, training costs keep rising, and companies keep looking for ways to avoid lock-in with any single cloud provider. Together is betting on that friction.