{"type":"post","id":"independent-bench-finds-mage-vl-cowork-asks-for-4341b9","slug":"independent-bench-finds-mage-vl-cowork-asks-for-4341b9","title":"Independent bench finds Mage-VL-Cowork asks for screenshots and never clicks outside Microsoft's harness","description":"Efrain Garay tested Microsoft’s Mage-VL-Cowork 4B on a 16 GB RTX 4070 Ti SUPER and found that it fit with usable response speed but never clicked a requested button in 30 trials.","retrieval_nugget":"Efrain Garay tested Microsoft’s Mage-VL-Cowork 4B on a 16 GB RTX 4070 Ti SUPER and found that it fit with usable response speed but never clicked a requested button in 30 trials. The result matters because it separates performance inside a computer-use harness from behavior a builder may encounter when the interaction loop changes.","published_at":"2026-11-14","updated_at":"2026-11-14","record_date":"2026-11-14","date_kind":"published_at","topics":["computer-use","evals"],"entities":["Microsoft","Mage-VL-Cowork","Mage-ViT","Qwen3-4B-Instruct-2507","Efrain Garay","OSWorld-Verified"],"related_patterns":["verification-bandwidth-is-the-scarce-resource"],"novelty_label":"new_fact","what_changed":"An independent RTX 4070 Ti SUPER test found Mage-VL-Cowork used 13,985 MB of VRAM and replied in 2.47 seconds, but failed to click in all 30 prompts, repeatedly asking for another screenshot.","prior_refs":["https://newruntime.com/posts/microsoft-releases-mage-vl-cowork-a-4b-hybrid-29e1a4/","https://newruntime.com/posts/microsoft-releases-mage-vl-cowork-a-4b-hybrid-111128/"],"source_url":"https://efraingaray.com/en/blog/mage-vl-cowork-16gb","source_urls":["https://efraingaray.com/en/blog/mage-vl-cowork-16gb"],"schema_version":"newruntime-agent-readable-v0.2","stable_id":"post:independent-bench-finds-mage-vl-cowork-asks-for-4341b9","status":"published","visuals":[],"editor":{"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]},"routes":{"html":"https://newruntime.com/posts/independent-bench-finds-mage-vl-cowork-asks-for-4341b9/","markdown":"https://newruntime.com/posts/independent-bench-finds-mage-vl-cowork-asks-for-4341b9.md","json":"https://newruntime.com/posts/independent-bench-finds-mage-vl-cowork-asks-for-4341b9.json"},"markdown_url":"https://newruntime.com/posts/independent-bench-finds-mage-vl-cowork-asks-for-4341b9.md","next_reads":[{"type":"topic","path":"/topics/computer-use/","url":"https://newruntime.com/topics/computer-use/","title":"Computer Use","reason":"Explore the computer-use topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/computer-use.md","json_url":"https://newruntime.com/topics/computer-use.json"},{"type":"topic","path":"/topics/evals/","url":"https://newruntime.com/topics/evals/","title":"Evals","reason":"Explore the evals topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/evals.md","json_url":"https://newruntime.com/topics/evals.json"},{"type":"collection","path":"/posts/","url":"https://newruntime.com/posts/","title":"Field Notes","reason":"Open the latest editorial analysis.","media_type":"text/html","json_url":"https://newruntime.com/posts.json"},{"type":"collection","path":"/sources/","url":"https://newruntime.com/sources/","title":"Sources","reason":"See the original sources this archive cites.","media_type":"text/html"}]}