- Together AI raised an $800 million Series C at an $8.3 billion post-money valuation, led by Saudi Aramco’s venture arm, with NVIDIA, Vista Equity Partners, and General Catalyst joining.
- The pitch: run open models like DeepSeek and Kimi at 6x to 60x lower cost than closed frontier APIs, protecting the margins that OpenAI-style pricing quietly erodes.
- Annual bookings crossed $1.15 billion last quarter, and the company plans to scale its compute footprint roughly 50-fold over the next five years.
- Aramco and Schneider Electric backing the round is a signal that energy giants now see AI infrastructure, not models, as the defining capital project of the decade.
On July 1, 2026, the venture arm of the world’s largest oil company led an $800 million bet not on a flashy chatbot, but on making artificial intelligence cheap. Together AI, a San Francisco startup that runs open-source models for enterprises, closed a Series C at an $8.3 billion valuation. The lead investor was Aramco Ventures. The subtext was unmistakable: the smart money has stopped chasing the next frontier model and started buying the infrastructure that makes every other model affordable.
The $800 Million Bet Against Closed AI
Who actually wrote the checks
Aramco Ventures led the round, but the cap table reads like a cross-section of who controls the AI economy. NVIDIA, Vista Equity Partners, General Catalyst, Emergence Capital, March Capital, Pegatron, and SentinelOne’s S Ventures all participated, alongside Schneider Electric’s venture fund. The valuation jump is steep: Together AI raised a $305 million Series B in February 2025 at roughly $3 billion, meaning its price tag has nearly tripled in under 18 months. For a company founded in 2022, that is a vertical climb.
What makes the round unusual is not the size but the identity of the lead. When an oil major’s investment arm writes the biggest check in an AI infrastructure deal, it is placing a long-duration bet that compute, not crude, is the resource worth cornering next.
Business Insight — Strategic investors rarely lead purely for financial return. Aramco taking the lead position signals that sovereign and energy capital view AI infrastructure as the next pipeline to own, and they would rather fund the demand for power than merely sell it.
Why Open Source Became a Margin Story
Frontier pricing eats the P&L
Together AI’s growth rests on an uncomfortable truth for any business built on closed-model APIs: frontier pricing can consume an entire product margin. The company positions itself as the infrastructure layer for the alternative, helping enterprises train and run open models such as DeepSeek, Nemotron, MiniMax, and Kimi at a fraction of the cost of closed systems, with comparable or better performance. Customers report savings of 6x to 60x versus closed-model pricing. AI customer-support startup Decagon, for example, cut its inference costs sixfold after moving to Together.
The demand data backs the thesis. Open-source model usage across the industry tripled over the past twelve months, and McKinsey found that nearly three-quarters of organizations expect to increase their use of open-source AI. Together AI’s own annual bookings crossed $1.15 billion last quarter, with paying customers that include Cursor, Cognition, and Decagon.
Business Insight — For most companies, the AI decision has quietly shifted from capability to unit economics. Once open models reach parity, the buying question becomes cost per token at scale, and that is where a 6x to 60x gap turns into a durable competitive moat.
Energy Money Meets Compute
Intelligence as cheap as electricity
Co-founder and CEO Vipul Ved Prakash framed the mission in terms an oil investor understands: “Intelligence is becoming a foundational resource for the modern economy, every bit as essential as electricity, bandwidth or capital. Our mission is to ensure that intelligence is abundant, not expensive.” Abhishek Shukla of Aramco’s Prosperity7 Ventures went further, calling AI infrastructure over the next decade “the biggest infrastructure project in human history.”
Schneider Electric’s participation makes the convergence literal. Its CEO noted that AI and energy infrastructure are merging, with efficiency as the link between them, because more efficient AI means less energy consumed per workload. Together AI says it will use the capital to scale its capacity and infrastructure footprint roughly 50-fold over the next five years, backed by the kind of long-term power and capital commitments only strategic investors can supply.
Business Insight — The AI buildout is increasingly an energy story wearing a software label. Whoever controls efficient compute controls the cost of intelligence, and energy incumbents have both the capital and the electrons to make that bet at a scale venture funds cannot match.
The Neocloud Race and Its Risk
Can a startup out-scale the hyperscalers?
Together AI is not alone in the so-called neocloud wave. Rivals like Groq and RunPod rent out raw GPU capacity, while TensorWave recently raised $350 million building on AMD chips and Upscale AI extended its round to a $2 billion valuation. What sets Together apart is that it bundles compute with its own inference-optimization software, which it says can cut the cost of running popular models by up to 80 percent. That software layer is the moat around what would otherwise be a commodity hardware business.
The risk is equally clear. The hyperscalers, flush with hundreds of billions in capital spending, are building their own inference capabilities at a scale no startup can match. Together AI’s wager is that open-source flexibility plus a software edge can defend its position before the giants catch up. The $800 million, and the identity of the backers, buys it time to find out.
Business Insight — In infrastructure markets, distribution and capital usually beat cleverness over the long run. Together’s survival depends on locking in enterprise switching costs through software and savings before hyperscalers commoditize the layer beneath it.
Related
- AI’s Real Money Isn’t in Models. Ask This $13B Startup.
- AI’s Next Crisis Isn’t Compute. It’s $570 Billion in Debt.
- The AI Lab That Refused Money Just Took $7.4 Billion
- Enterprises Bought $401B in GPUs. 95% Sit Idle.
- Oracle Booked $638 Billion. Then Its Stock Crashed.
Sources
- Together AI Raises $800 Million at $8.3 Billion Valuation to Make Frontier AI Accessible to All (Business Wire, July 1, 2026)
- Together AI raises $800 million in Series C led by Aramco Ventures (The Next Web, July 1, 2026)
- Neocloud Together AI raises $800M, leaps to $8.3B valuation (TechCrunch, July 1, 2026)
- Together AI raises $800 million Series C at $8.3 billion valuation (Quartz, July 1, 2026)
AI Biz Insider · AI Business EN · aibizinsider.com
댓글 남기기