landing-ai/vision-agent
An AI agent that takes image/video prompts and automatically generates runnable vision code to build visual AI apps.
Velocity · 7d
+6.3
★ / day
Trend
→steady
star history
VisionAgent is an agent framework from LandingAI that handles visual reasoning tasks. Given a prompt and an image or video, it automatically selects appropriate vision models and generates ready-to-run Python code. The system is designed to let developers build vision-enabled applications quickly by abstracting away model selection and code writing.