> This page is for version Dev.
> For other versions, use one of these documentation indexes:
> - 3.3.0 (default): https://docs.nvidia.com/vss/3.3.0/llms.txt
> - Dev: https://docs.nvidia.com/vss/dev/llms.txt

> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/vss/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/vss/_mcp/server.

# Performance

This section documents benchmark results for VSS agent workflows and Real-Time Video
Intelligence microservices. The results include measurements from VSS 3.2 and
3.3; each linked page identifies its own tested release, model, precision, and
hardware. Use those page-specific test configurations when sizing GPU
infrastructure or comparing use cases.

## Agent Workflows

| Agent Workflow      | Page                                                                                                   | What Is Measured                                                                    |                                                                                                     |                                                                                 |
| ------------------- | ------------------------------------------------------------------------------------------------------ | ----------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------- |
| Alert Verification  | [performance-alert-verification](/appendix/performance/agent-workflows/alert-verification-performance) | - End-to-end latency and verified alert throughput for live RTSP alert verification | RT-DETR and GDINO with CR3 across RTX PRO 4500, RTX PRO 6000, DGX H100 SXM, DGX Spark, and AGX Thor | VLM precision (BF16, FP8, NVFP4) and stream-concurrency scaling characteristics |
| Video Summarization | [performance-lvs](/appendix/performance/agent-workflows/video-summarization-performance)               | - E2E latency vs. video length (1 min–12 hr)                                        | Burst max concurrency at target avg latency (1, 5, 10 min videos)                                   | Cosmos3 Nano Reasoner FP8 across H100, RTX Pro 6000 SE, and L40S                |
| Search              | [performance-search](/appendix/performance/agent-workflows/search-performance)                         | - E2E mean and P90 latency vs. concurrent query load (1–50)                         | Per-stage latency breakdown on H100 and RTX Pro 6000 SE                                             | Video ingestion latency                                                         |

## Microservices

| Microservice                       | Page                                                                             | What Is Measured                                                     |                                                                     |                                                                       |                                                  |                                                                |
| ---------------------------------- | -------------------------------------------------------------------------------- | -------------------------------------------------------------------- | ------------------------------------------------------------------- | --------------------------------------------------------------------- | ------------------------------------------------ | -------------------------------------------------------------- |
| Real-Time CV (RT-CV)               | [performance-rt-cv](/appendix/performance/microservices/rt-cv-performance)       | - Maximum concurrent 1080p streams per GPU                           | RT-DETR (Resnet50 and EfficientViT/L2) and Grounding DINO models    | Latency-sensitive and throughput-maximizing operating modes           |                                                  |                                                                |
| Real-Time VLM (RT-VLM)             | [performance-rt-vlm](/appendix/performance/microservices/rt-vlm-performance)     | - Maximum concurrent streams and chunk latency for live RTSP streams | Total latency and throughput for pre-recorded video file processing | Alerting (OSL=1) and captioning (OSL=100) use cases                   | Cosmos3 Nano Reasoner via vLLM                   |                                                                |
| Real-Time Embedding (RT-Embedding) | [performance-rt-embed](/appendix/performance/microservices/rt-embed-performance) | - Maximum concurrent streams for live RTSP video embedding           | Video and text embedding throughput and latency in request mode     | File embedding latency vs. video length                               | Cosmos-Embed1, FP16, TensorRT                    |                                                                |
| VIOS                               | [performance-vios](/appendix/performance/microservices/vios-performance)         | - RTSP and WebRTC stream profiling (GPU/CPU utilization)             | Max concurrent streams per replica on L40 and H100                  | Single-stream and concurrent video-download latency (H.264 and H.265) | Single-stream and concurrent picture API latency | RTX PRO 6000 Blackwell benchmarks with and without transcoding |