Skip to main content
Back to trending

cosmo-edge

Production-grade C++ edge AI engine for video analytics and on-device VLM across Sophon, Rockchip RKNN, and x86, with visual orchestration, real-time OSD, events, and reproducible benchmarks.

  • bm1688
  • computer-vision
  • cpp17
  • cv186x
  • edge-ai
  • grounding-dino
  • on-device-ai
  • rk3576
  • rknn
  • rockchip
  • rv1126b
  • sophon
  • video-analytics
  • visual-pipeline
  • vlm
View on GitHub
Stars
1,548
Forks
267
+ today
+3
Created
4mo

Ranking data as of October 12, 2026 (UTC).

Star History

Today, hour by hour

1,548 stars
03

Overview

Production-grade C++ edge AI engine for video analytics and on-device VLM across Sophon, Rockchip RKNN, and x86, with visual orchestration, real-time OSD, events, and reproducible benchmarks. It ranks #107 on GitTiger, gaining +3 stars on October 12, 2026 (UTC).

The project is written in C and has 267 forks. It was created 4mo ago.

AI-generated summary of the repository README — may contain errors.

What it is

`cosmo-wander-ai/cosmo-edge` describes a C++ edge-AI engine for video analytics and on-device vision-language models. The repository description calls it production-grade and names Sophon, Rockchip RKNN, and x86 as target platforms.

Capabilities

The description lists visual orchestration, real-time on-screen display (OSD), event handling, and reproducible benchmarks. The tags also reference C++17, BM1688, CV186x, and Grounding DINO. These are supplied project claims and metadata; the available evidence does not include benchmark results, a device compatibility matrix, or deployment examples.

Getting started

No commands or prerequisites are included in the supplied fields. Check the repository README for supported hardware and models, build requirements, deployment guidance, and benchmark methodology before choosing a target device.

Installation
git clone https://github.com/cosmo-wander-ai/cosmo-edge.git
cd cosmo-edge
# see README for setup

Frequently Asked Questions

Which platforms are named?

Sophon, Rockchip RKNN, and x86 are named in the description.

What capabilities does the description list?

Video analytics and on-device VLM, visual orchestration, real-time OSD, events, and reproducible benchmarks.