DeepSeek rolls out image recognition mode for Chinese LLMs in beta test
On April 29, 2026, DeepSeek activated its "Image Recognition Mode" in a grey test release.
Published 29sem·1 source·updated 14sem
The fact
The feature enables analysis of image content, such as artifacts, charts, or food ingredients.
This deployment addresses a gap in Chinese language models, previously limited to text.
🔗Click the link to read an article on the topic:
What if you saw the whole news differently?Factae cross-checks hundreds of sources worldwide to keep only the fact, no opinion. Explore the front page.→Observed impact
Intensification de la concurrence des modèles d'IA entre acteurs chinois et occidentaux
Chine · © OpenStreetMap, OpenMapTiles, OpenFreeMapWorld map →Auto-synthesis from 1 media source · identified on May 2, 2026