Modal Raises $355M in Series C at $4.65B Valuation, Demonstrates Strong AI Infrastructure Traction
Key Takeaways
- ▸Modal raised $355M in Series C at $4.65B valuation, with General Catalyst and Redpoint leading the round
- ▸Company achieved 5x growth since September with $300M+ annualized revenue, demonstrating strong market traction in AI infrastructure
- ▸Over 1 billion sandboxes have been launched on Modal, highlighting demand for secure agent execution environments
Summary
Modal, a cloud infrastructure platform built specifically for AI workloads, has raised $355 million in Series C funding at a $4.65 billion post-money valuation. The round was led by General Catalyst and Redpoint, with new investors Menlo, Bain Capital Ventures, and Accel joining. The funding comes after Modal achieved 5x growth since September and surpassed $300 million in annualized revenue, demonstrating strong market demand for purpose-built AI infrastructure.
Unlike traditional cloud platforms designed for web applications, Modal provides primitives tailored to AI: elastic compute, low-latency inference, dynamic agent runtimes, reinforcement learning, and batch processing at scale. The company has become integral to AI development workflows, with over 1 billion sandboxes launched for agent execution, enabling companies like DoorDash, Suno, and Physical Intelligence to build AI-native applications.
Modal's momentum reflects a broader shift in how organizations are adopting AI—moving from reliance on frontier model APIs to owning and fine-tuning their own models using open-weight alternatives from DeepSeek and Qwen. The company is investing heavily in low-latency inference, open-source inference engines (Flash Attention, vLLM, SGLang), and solving fundamental infrastructure challenges like 100x GPU cold start improvements and sub-second scaling from 0 to 1,000 GPUs.
- Modal is positioning itself as the infrastructure layer for an era where developers increasingly own and fine-tune their own models
- The company is doubling down on inference optimization and open-source contributions to vLLM, SGLang, and Flash Attention



