Fireworks AI closed a $1.505 billion Series D funding round on July 16, 2026, pushing the AI infrastructure company’s post-money valuation to $17.5 billion. The round was led by Atreides Management, Index Ventures, and TCV, with follow-on participation from NVIDIA, Lightspeed Venture Partners, and Evantic Capital. Bessemer Venture Partners, Menlo Ventures, Insight Partners, Lone Pine Capital, Ontario Teachers’ Pension Plan, and Quantum Capital also joined the round, alongside a broader group of investors including 20VC, Operator Collective, Prysm Capital, Original Capital, and TIME Ventures.
The raise comes as Fireworks AI reported annualized revenue exceeding $1 billion, a figure the company described as roughly five times its prior year’s level. Enterprise customers including Uber and Shopify use the platform for specialized AI workloads, and the company positioned its growth as driven by accelerating demand for AI infrastructure capable of training and serving custom models at scale.
Fireworks AI focuses on specialized intelligence rather than general-purpose foundation models. Its core offering helps developers train and run custom AI models instead of relying on proprietary model APIs from large AI laboratories. The platform is built around speed-optimized inference infrastructure, developed by engineers with PyTorch experience, and is marketed as enabling cheaper and more controllable AI deployments for enterprises.
Fireworks AI builds for customization, not commoditization — giving developers control over the models they actually run.
The company also supports open-source AI models through cloud-based serving infrastructure, which it has positioned as an alternative to renting compute from larger model providers. The platform currently processes more than 10 trillion tokens daily, reflecting the scale at which enterprise workloads are being served across its infrastructure.
Proceeds from the Series D will fund expansion of global compute capacity, scaling of engineering teams, and a broader international infrastructure footprint to meet enterprise AI demand. The company also intends to deepen partnerships with Microsoft and NVIDIA as part of its growth strategy. A stated priority is strengthening its ability to serve open-source and custom models at lower cost than traditional model API alternatives.
The Series D represents a sharp acceleration in both valuation and capital raised. Fireworks AI’s prior Series C, completed in October 2025, raised $250 million at a $4 billion valuation. Before that, the company raised approximately $52 million in a mid-2024 Series B at a reported $552 million valuation and roughly $25 million in a Series A. Total funding before the Series D stood at approximately $327 million across all earlier rounds.
The $1.505 billion Series D alone exceeds 4.5 times the cumulative capital raised in all prior rounds combined.
The valuation jump from $4 billion to $17.5 billion between the Series C and Series D reflects a more than fourfold increase in roughly nine months. That trajectory places Fireworks AI among the faster-growing infrastructure companies in the current AI investment cycle, with its revenue growth and enterprise traction cited as central to investor interest in the round.
