Fireworks AI has raised $250 million in a Series C funding round at a $4 billion valuation. Founded by the engineers behind PyTorch, the round was led by Lightspeed Venture Partners, Index Ventures, and Evantic, with participation from existing investors including Sequoia Capital, NVIDIA, AMD, MongoDB, and Databricks. This latest funding brings Fireworks AI’s total capital raised to over $327 million.
Unlike the early years when AI focused heavily on model training, the industry’s attention is shifting to AI inference. Running trained models in production to make real-time predictions and decisions based on new data.
Fireworks AI provides a high-performance platform designed specifically for AI inference workloads at scale, powering AI applications for major companies such as Uber, Shopify, Genspark, Retell AI, and GitLab.
It delivers extremely fast and cost-effective inference using proprietary serverless infrastructure, custom CUDA kernels, advanced model sharding, and semantic caching technologies.
The platform achieves up to 12 times faster inference than leading benchmarks like vLLM and 40 times faster than GPT-4, resulting in significant operational cost savings for its enterprise customers.
An Accessible AI Infrastructure
The key thing that sets Fireworks AI apart is its embrace of open-source models across different types of data like text, images, audio, and even combinations of them. Unlike many black-box AI providers, Fireworks gives enterprises greater transparency and control, building an ecosystem that’s user-friendly for developers and scalable for mission-critical enterprise use.
The platform supports over 100 state-of-the-art AI models and offers powerful tools for fine-tuning and optimizing model performance through techniques like reinforcement learning, quantization-aware tuning, and adaptive speculation.
With simple APIs and on-demand, auto-scaling GPU resources, Fireworks AI makes it easy for organizations to move rapidly from prototype to production while retaining full ownership of their AI models and data.
Even though Fireworks AI is gaining rapid traction, the company faces the challenge of widespread enterprise AI adoption. Many organizations are still building internal expertise to develop and deploy AI models tailored to their unique business needs. Gartner analyst Chirag Dekate notes that roughly 80% of enterprises have yet to reach advanced AI engineering maturity.
The company’s growth reflects this market opportunity. Fireworks AI reported $280 million in annual recurring revenue as of May 2025, representing explosive growth from $6.5 million a year prior. Even its developer base nearly doubled in 2024.
Market Position
The company’s CEO and co-founder, Lin Qiao, was part of the team that originally created PyTorch at Meta, one of the most influential open-source AI frameworks globally. Using this deep expertise, Fireworks AI aims to empower enterprises with automated product and model co-design capabilities, enabling them to achieve unparalleled speed, quality, and cost-efficiency in generative AI.
“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 Qiao. “Fireworks AI is the only platform delivering 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, and GitLab have scaled on Fireworks.”
Fireworks AI has forged partnerships with industry leaders like Google Cloud, NVIDIA’s Inception program, and MongoDB. The startup competes against both hyperscale cloud providers such as AWS, Google Vertex AI, and Microsoft Azure, as well as specialized AI infrastructure startups like Hugging Face, Groq, OpenRouter, and Replicate.
With $250 million in fresh capital, Fireworks AI plans to accelerate its expansion across global markets, scale engineering, product, and GTM teams while enhancing its hardware capacity to meet the growing demands of enterprise AI.
Fireworks AI’s focus on democratizing AI infrastructure promises to empower a broad spectrum of organizations, enabling them to innovate rapidly while maintaining ownership and control over their AI technologies.
