Qwen3-VLã®è»½éçã§ããQwen3-VL-4Bããã³Qwen3-VL-8Bãå ¬éããã¾ããããããã®ã¢ãã«ã¯ãQwen3-VLã®å ¨æ©è½ãä¿æããã¦ããã徿¥ã®ã¢ãã«ã«æ¯ã¹ã¦å¤§å¹ ã«VRAM使ç¨éã忏ãã¤ã¤ãQwen2.5-VL-72Bã«å¹æµããæ§è½ãçºæ®ããã¨ã®ãã¨ã§è©¦ãã«ãã¼ã«ã«PCã«ã¤ã³ã¹ãã¼ã«ãã¦ä½¿ã£ã¦ã¿ã¾ããã Qwen https://qwen.ai/home Introducing the compact, dense versions of Qwen3-VL â now available in 4B and 8B pairs, each with both Instruct and Thinking variants. â Lower VRAM usage â Full Qwen3-VL capabilities retained â Strong performan


{{#tags}}- {{label}}
{{/tags}}