Milestone Introduces VLM as a Service to Accelerate AI-Powered Video Applications
Milestone Systems has unveiled Hafnia Vision Language Model as a Service (VLMaaS), giving developers, integrators and technology partners API-first access to advanced, production-ready video intelligence powered by NVIDIA Cosmos Reason.
Designed to remove the complexity traditionally associated with AI development, VLMaaS enables teams to add generative AI video capabilities to existing applications without the need to build, fine-tune or manage their own models. The service supports everything from rapid MVP testing through to large-scale production deployments.
According to Milestone, VLMaaS can reduce development effort by up to 70 times compared with building and fine-tuning a vision language model independently. The model is optimised for real-world traffic video and supports prompt-based instructions for traffic-related operations, making it suitable for smart city, transport and infrastructure use cases.
The platform is delivered via HTTPS APIs, supports regional models for the US and EU, and can be used either as a standalone solution or integrated directly with Milestone’s broader product portfolio. Pricing follows a pay-per-use model based on API calls, avoiding large upfront costs or custom training investments.
Both VLMaaS and Milestone’s new Video Summarisation for XProtect are powered by the Hafnia VLM, which has been fine-tuned on 75,000 hours of responsibly sourced real-world video data from Europe and the United States. Data preparation is handled using NVIDIA Cosmos Curator, with deployment options across cloud infrastructure or regional data centres.
Andrew Burnett, Interim Chief Technology Officer at Milestone Systems, said the new offerings directly address video overload and manual processing bottlenecks.
“Operators get immediate insight directly within XProtect, while builders get API-first access to production-ready intelligence without bespoke training or heavy infrastructure,” Burnett said. “Because the model is specialised for real-world traffic video and trained on responsibly sourced data, customers can deploy with confidence and enhance the solutions they already have in place.”
Cities including Genoa, Italy, and Dubuque, Iowa, are already preparing to adopt the new capabilities as part of advanced traffic management initiatives.
