Next upHack for Humanity: San Francisco (powered by Google Gemini)
News

Together AI raises $800M Series C led by Aramco Ventures, doubling valuation to $8.3B

Together AI raised an $800 million Series C led by Aramco Ventures, more than doubling its valuation to $8.3 billion from $3.3 billion.

Dmytro Spodarets
Jul 1, 2026 · 2 min read

Together AI, the cloud platform for open-weight AI model training and inference, raised an $800 million Series C led by Aramco Ventures on July 1, 2026, more than doubling its valuation to $8.3 billion. The mark is up from the $3.3 billion set at its February 2025 Series B.

The round is a bet on scale. Together AI says it plans to expand computing capacity roughly 50-fold over five years, and that new investors are separately capitalizing more than 500 megawatts of additional compute, a signal of how much power the open-weight inference business now demands.

Founded in 2022 by Vipul Ved Prakash, Percy Liang and Ce Zhang, the company positions itself as an alternative to closed frontier labs, renting GPU capacity to developers who run open models rather than proprietary ones. It says annual bookings topped $1.15 billion last quarter, up sharply as demand for inference has grown.

Beyond Aramco Ventures, the round drew Vista Equity Partners, General Catalyst, Emergence Capital, Nvidia, Salesforce Ventures, March Capital, Pegatron and SentinelOne’s S Ventures. Together AI has now raised well past its earlier $305 million Series B and $102.5 million Series A.

The company cites customers including Cognition, Decagon, Eleven Labs, Cursor and Suno, and says Decagon reduced its inference costs sixfold after moving to the platform. That figure is the company’s own; it has not been independently verified, and customer-cost claims from infrastructure vendors typically reflect specific workloads rather than across-the-board savings.

The $8.3 billion mark places Together AI among the more richly valued of the “neocloud” providers competing to supply GPU capacity outside the hyperscalers. Whether the 50-fold capacity plan holds will depend on power, chips and the durability of open-weight demand as frontier labs push their own hosted inference. For now, the money says investors expect that demand to keep climbing.


Dmytro Spodarets
Dmytro Spodarets
Founder & Editor-in-Chief

Founder and Chief Editor of Data Phoenix — a San Francisco Bay Area media and education platform focused on AI and Data.

More news