Together AI announced an $800 million Series C on Wednesday at an $8.3 billion post-money valuation, more than doubling the San Francisco company's prior valuation. The round was led by Aramco Ventures, the venture arm of Saudi Arabia's state oil company Saudi Aramco, with participation from Vista Equity Partners, General Catalyst, Emergence Capital, Nvidia, March Capital, Pegatron, SE Ventures, and others. Together AI's platform gives businesses access to open-source AI models – including DeepSeek, MiniMax, and Kimi – at significantly lower cost than closed systems from OpenAI and Anthropic. The company acquires GPU compute through purchase or rental, deploys inference-optimisation software, and makes hundreds of open-source models available in configurations that can be customised to client specifications. Founded in 2022, it counts Cursor, Cognition, and Decagon among its paying customers. Decagon co-founder Ashwin Sreenivas said that moving workloads to Together AI cost between one-fifth and one-seventh of what closed models charged. KeyToFinancialTrends opens on that cost differential as the single number that explains why this raise closed at $8.3 billion: 5-7x cheaper inference is not a marginal efficiency gain – it is the kind of spread that restructures procurement decisions at the enterprise level.
The revenue data accompanying the raise is the more important signal. Together AI's annual bookings crossed $1.15 billion last quarter, the company said, as open-source model usage across the industry tripled in twelve months. Those two data points together – $1.15 billion in bookings and industry-wide open-source adoption tripling – tell a story about the structural shift underway in AI procurement rather than just Together AI's specific performance. Enterprise buyers who spent 2024 and early 2025 anchored to GPT-4 and Claude for their AI workloads are actively diversifying to open-source alternatives where the economics are better and the data sovereignty argument – raised pointedly by Palantir CEO Alex Karp in remarks on the same day as this funding announcement – is easier to satisfy. KeyToFinancialTrends positions Together AI's $1.15 billion bookings as the most concrete evidence yet that the open-source migration is beyond the pilot stage.
The Aramco Ventures lead is strategically notable. Saudi Arabia's sovereign AI investment vehicle Humain has been the most visible Saudi AI deployment entity, but Aramco Ventures represents a different funding philosophy – it invests to serve Saudi Aramco's operational needs as well as to generate returns. An energy company of Aramco's scale processes enormous volumes of geological, operational, and logistics data that are ideal candidates for AI workloads. The Together AI platform – open-source models deployable on Aramco's own infrastructure or through Together's cloud – aligns directly with a large industrial enterprise that wants AI capabilities without transferring proprietary operational data to third-party model providers. Analysts at KeyToFinancialTrends read the Aramco Ventures lead not just as Gulf capital seeking AI exposure but as an anchor customer relationship embedded in a funding round.
Position Together AI's valuation against its competitive set. Baseten, a direct competitor in the AI inference infrastructure market, secured $1.5 billion in financing at a $13 billion valuation just last month. That gives Baseten a higher absolute valuation but a lower revenue multiple than Together AI, whose $1.15 billion in annual bookings against an $8.3 billion valuation implies a roughly 7x revenue multiple. Both companies are growing fast enough that any static multiple comparison is quickly outdated. The more useful comparison is market positioning: Together AI's emphasis on open-source model diversity – hundreds of models available for customisation – differentiates it from Baseten's more infrastructure-focused approach, and the open-source angle aligns directly with the enterprise customer concern about vendor dependency that Karp identified as a widespread problem.
The IPO signal embedded in the fundraise is worth noting without overinterpreting it. Prosperity7 Ventures managing director Abhishek Shukla said Together AI is headed toward the public markets. That is a forward-looking statement from an investor, not a management commitment, and the timeline is unspecified. The raise brings Together AI's total funding to $1.3 billion. A company with $1.15 billion in annual bookings growing at the pace the company implies would have a reasonable case for public market access within two years, but the AI IPO window has been selective and conditions can change.
The infrastructure ambition in the use-of-proceeds statement is worth a paragraph on its own. Together AI said it plans to grow its computing capacity by approximately 50 times over the next five years. At current scale, that implies an infrastructure buildout that would require continued capital access well beyond this $800 million round. The GPU supply chain, data centre power procurement, and networking costs at 50x current capacity are not trivial to finance or execute. Key To Financial Trends sizes up this infrastructure commitment as the most operationally ambitious statement in the announcement, and the one most likely to require a follow-on raise or a partnership arrangement with a hyperscaler before the five-year target is reached.
