Decart, which offers real-time generative video and GPU optimization tech to cloud providers and AI companies, raised a $100M Series B at a $3.1B valuation
and the future of real-time creative AI just got a whole lot closer. … Decart : We've raised $100M to scale real-time, interactive AI. — Our Live Stream Diffusion is proof of what's possible - a new category of real …
Context & Ripple Effects
Decart had previously raised a $32M Series A to support AI model training at scale, positioning its newer funding round as an escalation from training infrastructure into real-time generative video and GPU optimization.
The company sells to cloud providers and AI companies, so the round matters beyond a single creative-AI application: it funds technology intended to make interactive generation more practical on deployed GPU infrastructure.
First-order effects
- Decart gains $100M to scale its real-time, interactive AI offering, with a $3.1B valuation giving it a substantial financing base for product development and customer deployment.
- Cloud-provider and AI-company customers have a better-capitalized vendor focused on both real-time video generation and GPU optimization.
Second-order effects
- Competing real-time generative-AI vendors face greater pressure to demonstrate not just model quality but latency, interactivity and efficient GPU use; Cartesia's real-time voice-AI funding reflects the same performance-oriented competitive field.
- For cloud customers, optimization software becomes more strategically tied to generative-video rollouts, because model capability and the cost of serving it are increasingly inseparable.
Third-order effects
- If well-funded vendors can turn real-time generation into a dependable cloud workload, competitive advantage may shift toward companies that pair models with efficient serving infrastructure rather than offering generation alone.
- The financing points to a broader separation between experimental creative AI and production-grade interactive systems, though adoption will depend on whether providers can deliver the claimed performance at viable operating costs.
The trend: Generative AI investment is moving toward low-latency, interactive products whose commercial value depends as much on compute efficiency as on model output.