kslays I have a sample image and want to know the font used. With web access off, I asked Kimi K2 and it appeared to not have access to the image. Kimi K2 no web: https://kagi.com/assistant/f181652c-485f-4c0f-9967-dab1a0aa2f09 Kimi K2 with web: https://kagi.com/assistant/fa45edbc-582a-416e-8825-7385cdc94c42 Qwen 235B (fast) partially worked: https://kagi.com/assistant/d1333fde-c4d0-4a79-910e-bd882f94b014 Ki appeared to work: https://kagi.com/assistant/8685aab6-009f-4402-91c9-796b16b3ac8a The assistant should be able to analyze the image for font content, with or without web access. If a model does not have vision, the Assistant should make that clear with an error message before the LLM responds. (The screenshot was taken from this web page: https://www.nationalparks.nsw.gov.au/things-to-do/experiences/gidjuum-gulganyi-walk )
yiwei-1 Kimi K2 does not have vision support, so it cannot natively analyze images to identify fonts (they are OCR’d first). We will update the model info box to better indicate which models can process images natively.
