Back to News
Alibaba (ModelScope / Qwen)•May 9, 2026
Qwen3.6 Instruction-Oriented Object Detection
Paid
Object Detection
Computer Vision
Qwen
ModelScope
Multimodal AI
### TL;DR
ModelScope showcases a new Instruction-Oriented Object Detection capability powered by Qwen3.6-35B-A3B. Using natural language instructions, the model can detect and localize any objects in images, returning precise bounding boxes and structured JSON outputs — a major leap in vision-language understanding.
Key Insights & Metrics
Pricing
Free demo on ModelScope
Cost structure
Version
Qwen3.6-35B-A3B
Current release version
Hardware
Cloud API / Demo
Compute requirements
Category
Paid
Licensing model
Region
China
Primary region
Key Features
- Natural language instruction-driven object detection using Qwen3.6-35B-A3B
- Generates precise bounding boxes and structured JSON outputs from text descriptions
- Available as a live interactive demo on ModelScope platform
Discussion
0
Upvotes
0
Downvotes
0 reviews
Sign in to leave a review
Reviews
No reviews yet. Be the first to review!