Investors Lightspeed, Index Ventures and Evantic bet on Fireworks as the infrastructure backbone for enterprise AI adoption.

REDWOOD CITY, Calif.: Fireworks AI, the AI inference cloud powering production AI applications for companies like Uber, Genspark, and Shopify, today announced a $250 million Series C funding round at a $4 billion valuation. The round, which includes primary and secondary funding, was led by Lightspeed Venture Partners, Index Ventures, and Evantic, with participation from existing investor Sequoia Capital. This brings Fireworks’ total funding to $327 million.

AI is no longer an experiment; it is the battlefield for competitiveness. Enterprises can’t afford to hand over their data, products, and customer experience to generic platforms that trade cost for lock-in. Fireworks was built to put control back in the hands of builders.

“Our mission is to enable every business to achieve automated product and model co-design to reach maximum quality, speed, and cost-efficiency using generative AI,” said Lin Qiao, CEO and co-founder of Fireworks. “Fireworks is the only platform that delivers all three: state-of-the-art open-source models, sub-second inference at scale, and the ability to own and differentiate your AI. That’s why companies like Uber, Genspark, Retell AI, Shopify, GitLab, and so many AI Native startups have built and scaled on Fireworks.”

Generic models can’t capture the valuable data that lives inside enterprises and their applications, including user interactions, domain-specific workflows, proprietary knowledge bases, and behavioral patterns. Fireworks helps developers tune their models for specific use cases using this enterprise- and application-specific data. And as customers interact with these customized applications, new data continuously feeds back into and improves the system. Fireworks’ product-model co-design approach combines a tight data feedback loop with continuous evaluation and reinforcement learning that allows enterprises to sharpen their AI applications over time, optimizing for cost, speed, and quality simultaneously.

Founded by the team behind PyTorch, Fireworks scaled to process more than 10 trillion tokens every single day for over 10,000 customers in 2025. This year also saw the launch of Eval Protocol, the first serious attempt to bring order to the chaos of model evaluation, and application-tailored tuning (e.g., reinforcement learning), the first tool that lets developers train open-source models with the same playbook frontier labs guard so closely.

While others hand out black-box APIs, Fireworks arms teams with access to hundreds of cutting-edge open-source models across text, image, audio, and multimodal domains, with fine-tuning options that range from simple supervised training to reinforcement learning.

“We are really excited to see what we were able to achieve partnering with Fireworks,” said Kay Zhu, co-founder and CTO at Genspark. “Fireworks enabled us to own our AI journey, and unlock better quality in just four weeks. This resulted in a better user experience for our customers.”

The new funding will accelerate Fireworks’ campaign to reshape AI infrastructure and meet skyrocketing enterprise demand. It will go toward global expansion, aggressive investment in cutting-edge algorithms, advancing model quality, software infrastructure, developer toolchains, and hiring across engineering, product, and GTM roles. While others double down on lock-in, Fireworks is doubling down on developers-giving them the power to build, deploy, and scale AI on their own terms.

About Fireworks AI

Fireworks AI is the global AI inference cloud and infrastructure platform that enables teams like Cursor, Uber, DoorDash, and Shopify to build, tune, and scale highly optimized generative AI applications. Founded by the team behind PyTorch, and backed by Sequoia, Benchmark, Lightspeed, Index, Nvidia, and AMD. Fireworks provides deep support for hundreds of state-of-the-art open models in text, image, audio, embedding, and multi-modal formats globally. Visit https://fireworks.ai/ for more information.

Fonte: Business Wire