Together AI‘s Series C has landed at $800 million, pushing the open-source inference platform to an $8.3 billion post-money valuation, the company announced via Business Wire. That is the kind of number that makes you sit up, especially for a company that only existed on paper four years ago.
Aramco’s Prosperity7 Leads the Together AI Series C
The round was led by Aramco Ventures, specifically through its Prosperity7 Ventures arm, the diversified venturing programme of Saudi Aramco’s investment vehicle. That is not a first encounter: The Next Web reports that Prosperity7 co-led Together AI’s Series B alongside General Catalyst, so the relationship predates this cheque.
Other participants in the Series C include Vista Equity Partners, General Catalyst, Emergence Capital, Nvidia, March Capital, Pegatron, SentinelOne’s S Ventures, and, per Reuters, Salesforce Ventures, a name absent from Together AI’s own investor list in the press release. The omission is worth a raised eyebrow but not a conspiracy theory.
From $3.3bn to $8.3bn in 16 Months
The fundraising trajectory is worth spelling out. Together AI raised a $305 million Series B at a $3.3 billion valuation roughly 16 months ago, itself preceded by a $102.5 million Series A led by Kleiner Perkins in 2023. This latest close brings total funding raised to more than $1.3 billion, according to Tech Funding News.
Back in March, The Information reported that the company was seeking $1 billion at a $7.5 billion valuation. The final terms show Together AI took less money and landed at a higher valuation than those early whispers suggested, not a bad outcome, assuming you trust the maths on a pre-revenue infrastructure bet of this scale.
CEO Vipul Ved Prakash says the platform has become ‘one of the largest producers of AI tokens in the world,’ with the company generating roughly $1 billion in annualised revenue at the time of the announcement. The company also reported annual bookings of over $1.15 billion as of its last quarter. Those two figures are not identical things (bookings and revenue are different animals) but together they paint a picture of a business that is moving real money.
Open Source Is Eating the AI Budget
The pitch underneath all of this is straightforward: enterprises are quietly swapping expensive closed frontier models for cheaper open-source alternatives. Together AI’s platform supports models including DeepSeek, Nemotron, MiniMax, and Kimi, giving customers a broad menu without the per-token premium that comes with calling a closed API.
The company says open-source model usage has tripled across the industry over the past year, citing research from AI gateway OpenRouter. Paying customers named publicly include Cursor, Cognition, and Decagon, a shortlist that skews heavily developer, which is where open-source adoption tends to start before it spreads into enterprise procurement.
The new capital is earmarked to scale Together AI’s infrastructure footprint approximately 50-fold over the next five years. Whether the market holds together long enough to justify a 50x build-out is the sort of question the round’s press materials politely decline to address.
A Crowded Lane Is Getting More Expensive
The neocloud investment frenzy is not limited to Together AI. Upscale AI raised a combined $500 million across a Series A and extension at a $2 billion valuation last month. TensorWave, which bets on AMD GPU clusters rather than Nvidia’s, pulled in $350 million at a $1.55 billion valuation in the same period. Fireworks AI, a direct rival in open-source inference, raised $250 million at a $4 billion valuation in October 2025.
Then there is Groq, which licensed its chip technology to Nvidia for around $20 billion in December 2025, then turned around and raised $650 million in June 2026 to rebuild as an AI inference cloud provider. The inference infrastructure market is, in short, awash with capital and short on settled winners.
Together AI was co-founded by Prakash, who previously sold social media search platform Topsy to Apple in 2013 for a reported $200 million-plus, alongside Stanford professor Percy Liang and Ce Zhang, an associate professor at ETH Zürich and the University of Chicago who serves as the company’s CTO.
The real test arrives when enterprises stop comparing open-source token costs to frontier model prices and start comparing them to each other. At that point, infrastructure differentiation will matter a lot more than a good fundraising story.
