Testing Qwen 3 Image: Map Borders Shift Based on Prompts, Includes Chinese Labels
NirantK · x · 2026-07-22
NirantK conducted a boundary test on the Qwen 3 Image generation model. The results show that when asked for a map of India, the model includes Pakistan-administered Kashmir; conversely, when asked for a map of Pakistan, it includes the complete Kashmir region.
Furthermore, even when using English prompts, the generated map images contain Chinese text labels.
Related event: Qwen Image Model Sparks Controversy Over India-Pakistan Map Generation(2 posts)→
More from Multimodal
- Storyboard-first workflows are making AI dance videos and influencers more consistent — aftahi_ai · 2026-07-22
- Interactive video should be judged by responsiveness, not just frame quality — Soggy_Limit8864 · 2026-07-22
- Runpod MCP and Claude help spin up image and video generation workflows — 802high · 2026-07-22
- Midjourney prompt turns a bee into a glitching pixel explosion — michaelrabone · 2026-07-22
- A physics reward can improve video generation without creating a real physics engine — Dapper-Drawer4546 · 2026-07-22
- HeyGen adds a media-sourcing skill for coding agents with 75k images and 10k tracks — HeyGen · 2026-07-22