Archive — CN — technode
DeepSeek begins limited testing of vision mode
On April 29, DeepSeek commenced limited testing of its vision mode, available to select users on its web and app platforms. This new feature introduces multimodal capabilities alongside its existing Fast and Expert modes, enhancing its offerings beyond basic text recognition.
Key points
- DeepSeek began testing vision mode on April 29.
- Limited access granted to select users through web and app updates.
- Vision mode introduces multimodal capabilities beyond OCR.
- This feature complements existing Fast Mode and Expert Mode.
Why it matters
- The rise of multimodal AI tools is a current trend in IT and technology.
- Similar advances are seen in competitors focusing on diverse input methods.
- By 2025, the demand for enhanced AI capabilities is expected to increase, making technologies like DeepSeek's significant.
DeepSeek's vision mode marks an important step in evolving its AI capabilities, aligning with the broader IT industry shift towards multimodal technologies and enhancing user engagement.