--- title: Lower the Cost of Building and Running Visual AI Agents... description: NVIDIA's VSS Blueprint 3.3 lowers both development and runtime costs for building visual AI agents that process video with vision-language models. The Build... canonical: https://daily.dev/posts/lower-the-cost-of-building-and-running-visual-ai-agents-with-nvidia-vss-blueprint-3-3-m3dwypzn6 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Lower the Cost of Building and Running Visual AI Agents with NVIDIA VSS Blueprint 3.3 | daily.dev og:description: NVIDIA's VSS Blueprint 3.3 lowers both development and runtime costs for building visual AI agents that process video with vision-language models. The Build... og:url: https://daily.dev/posts/lower-the-cost-of-building-and-running-visual-ai-agents-with-nvidia-vss-blueprint-3-3-m3dwypzn6 og:image: https://api.daily.dev/og/posts/M3dWypzN6.png og:image:alt: Lower the Cost of Building and Running Visual AI Agents with NVIDIA VSS Blueprint 3.3 og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Lower the Cost of Building and Running Visual AI Agents with NVIDIA VSS Blueprint 3.3 **[NVIDIA Developer](https://daily.dev/sources/nvidiadev)** · 10 min read · 2 upvotes · 0 comments ## Summary NVIDIA's VSS Blueprint 3.3 lowers both development and runtime costs for building visual AI agents that process video with vision-language models. The Build Vision Agent skill (vss-build-vision-ai) lets developers describe an application in natural language and have a coding agent assemble a deployment from validated profiles, reusing shared infrastructure like Kafka, Redis, and Elasticsearch, cutting a bottling-line overflow agent build to under 30 minutes. On the runtime side, Adaptive Efficient Video Sampling (EVS) prunes unchanged visual tokens between frames and batches processing around events, cutting alert latency 17%, increasing concurrent real-time VLM streams 46%, and reducing VLM input tokens by 80% for a 60-minute video summary. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is Adaptive EVS in NVIDIA VSS 3.3 and how much does it reduce VLM processing cost? Adaptive EVS (Efficient Video Sampling) is a feature in NVIDIA VSS Blueprint 3.3 that prunes visual tokens for unchanged regions between video frames and batches processing around detected events. On an RTX PRO 6000 Blackwell GPU running Cosmos 3 Super FP8, it cut alert contextualization latency 17% (1,021ms to 844ms), increased concurrent real-time VLM streams 46% (13 to 19), and reduced VLM input tokens 80% for a 60-minute video summary. _Track releases like this on daily.dev to stay current on GPU-efficient video AI pipelines._ ### How does the Build Vision Agent skill in NVIDIA VSS 3.3 speed up building a video AI agent? The vss-build-vision-ai skill lets a developer describe a desired application in natural language, then it selects the closest of four validated profiles (base, alerts, lvs, search) as a Foundation and computes the smallest delta needed, reusing shared services like Kafka, Redis, and Elasticsearch instead of duplicating infrastructure. A recorded example built a bottling-line overflow detection agent with two RTSP cameras, alert verification, search, and reporting to a live, previewable deployment in under 30 minutes. _Developers composing multi-workflow AI agents can follow rollouts like this on daily.dev._ ### What VLM and LLM does NVIDIA VSS use for video search and summarization agents? NVIDIA VSS connects vision-language models such as NVIDIA Cosmos with LLMs such as NVIDIA Nemotron, along with retrieval-augmented generation (RAG) and Model Context Protocol (MCP) tools, to turn live and recorded video into natural-language search, visual Q&A, verified alerts, and automated reporting. _daily.dev helps engineers evaluating VLM and LLM stacks for video agents track ecosystem updates._ ## Similar posts on daily.dev - [Transform Video Into Instantly Searchable, Actionable Intelligence with AI Agents and Skills](https://daily.dev/posts/transform-video-into-instantly-searchable-actionable-intelligence-with-ai-agents-and-skills-o5xf3vza4) · NVIDIA Developer · 1 upvotes · 0 comments - [Make Sense of Video Analytics by Integrating NVIDIA AI Blueprints](https://daily.dev/posts/make-sense-of-video-analytics-by-integrating-nvidia-ai-blueprints-e6bohxhl5) · NVIDIA Developer · 0 upvotes · 0 comments - [Into the Omniverse: Three Workflows for Improving Vision AI Agent Accuracy With Synthetic Data and Fine-Tuning](https://daily.dev/posts/into-the-omniverse-three-workflows-for-improving-vision-ai-agent-accuracy-with-synthetic-data-and-f-4egob4ssz) · NVIDIA · 0 upvotes · 0 comments - [Run Highly Efficient and Accurate Multi-Agent AI with NVIDIA Nemotron 3 Super Using vLLM](https://daily.dev/posts/run-highly-efficient-and-accurate-multi-agent-ai-with-nvidia-nemotron-3-super-using-vllm-cpvrpiesj) · vLLM · 1 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#gpu](https://daily.dev/tags/gpu), [#vlm](https://daily.dev/tags/vlm) [View this post on daily.dev](https://daily.dev/posts/lower-the-cost-of-building-and-running-visual-ai-agents-with-nvidia-vss-blueprint-3-3-m3dwypzn6) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Lower the Cost of Building and Running Visual AI Agents with NVIDIA VSS Blueprint 3.3","url":"https://daily.dev/posts/lower-the-cost-of-building-and-running-visual-ai-agents-with-nvidia-vss-blueprint-3-3-m3dwypzn6","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/lower-the-cost-of-building-and-running-visual-ai-agents-with-nvidia-vss-blueprint-3-3-m3dwypzn6"},"datePublished":"2026-09-29T18:38:48.610Z","dateModified":"2026-09-29T18:39:15.690Z","description":"NVIDIA's VSS Blueprint 3.3 lowers both development and runtime costs for building visual AI agents that process video with vision-language models. The Build...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4a1927d1bac5685a31a1c1d78f12419d?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4a1927d1bac5685a31a1c1d78f12419d?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"NVIDIA Developer","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"NVIDIA Developer","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/86e45aab42ba48ce83103d01b1119910","url":"https://daily.dev/sources/nvidiadev"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/lower-the-cost-of-building-and-running-visual-ai-agents-with-nvidia-vss-blueprint-3-3-m3dwypzn6","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,gpu,vlm","timeRequired":"PT10M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"NVIDIA Developer","item":"https://daily.dev/sources/nvidiadev"},{"@type":"ListItem","position":3,"name":"Lower the Cost of Building and Running Visual AI Agents with NVIDIA VSS Blueprint 3.3"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/lower-the-cost-of-building-and-running-visual-ai-agents-with-nvidia-vss-blueprint-3-3-m3dwypzn6#faq","mainEntity":[{"@type":"Question","name":"What is Adaptive EVS in NVIDIA VSS 3.3 and how much does it reduce VLM processing cost?","acceptedAnswer":{"@type":"Answer","text":"Adaptive EVS (Efficient Video Sampling) is a feature in NVIDIA VSS Blueprint 3.3 that prunes visual tokens for unchanged regions between video frames and batches processing around detected events. On an RTX PRO 6000 Blackwell GPU running Cosmos 3 Super FP8, it cut alert contextualization latency 17% (1,021ms to 844ms), increased concurrent real-time VLM streams 46% (13 to 19), and reduced VLM input tokens 80% for a 60-minute video summary. Track releases like this on daily.dev to stay current on GPU-efficient video AI pipelines."}},{"@type":"Question","name":"How does the Build Vision Agent skill in NVIDIA VSS 3.3 speed up building a video AI agent?","acceptedAnswer":{"@type":"Answer","text":"The vss-build-vision-ai skill lets a developer describe a desired application in natural language, then it selects the closest of four validated profiles (base, alerts, lvs, search) as a Foundation and computes the smallest delta needed, reusing shared services like Kafka, Redis, and Elasticsearch instead of duplicating infrastructure. A recorded example built a bottling-line overflow detection agent with two RTSP cameras, alert verification, search, and reporting to a live, previewable deployment in under 30 minutes. Developers composing multi-workflow AI agents can follow rollouts like this on daily.dev."}},{"@type":"Question","name":"What VLM and LLM does NVIDIA VSS use for video search and summarization agents?","acceptedAnswer":{"@type":"Answer","text":"NVIDIA VSS connects vision-language models such as NVIDIA Cosmos with LLMs such as NVIDIA Nemotron, along with retrieval-augmented generation (RAG) and Model Context Protocol (MCP) tools, to turn live and recorded video into natural-language search, visual Q&A, verified alerts, and automated reporting. daily.dev helps engineers evaluating VLM and LLM stacks for video agents track ecosystem updates."}}]} ```