Qwen Image 2.1 launches: 7B image generation and editing model with up to 10 reference images
multimodalart · x · 2026-09-21
- Qwen Image 2.1 is out: a 7B-parameter native image generation and editing model supporting up to 10 image references.
- It ships with its own prompt-enhancement LLMs and is integrated with diffusers and ComfyUI.
- Available to try on Hugging Face Spaces.
More from Multimodal
- A Tree Through Four Seasons in a Bottle, Built With Three.js + TSL and Suno — techartist_ · 2026-09-21
- MiniMax H3 License Doesn't Apply in US, EU, UK, or South Korea — Read Before Installing — LeoLeg76 · 2026-09-21
- Picteus: open-source local indexer gives every Stable Diffusion output a searchable full recipe — shivam_dewan · 2026-09-21
- Nano Banana 2.5 image model expected next week, with thinking levels and up to 4K output — koltregaskes · 2026-09-21
- AI art community share: underrated AI artists and OG Tezos crypto-artists to follow — Merzmensch · 2026-09-21
- Early Qwen-Image 2.1 tests: art styles OK, realism 'so bad' tester regretted it — No_Daikon3851 · 2026-09-21