Proximal @ProximalHQ · 02 Sep 2026
GLM 5.3, the third-best model on FrontierSWE and strongest open source model, does not have native vision capabilities (we are currently testing GLM-5.3 Flash, which has vision support) In visual tasks, we observe GLM converting images to ASCII art to read their contents https://t.co/vYlYXgaTCH
6 201Views
63Likes
3Reposts
1Replies
1Quotes
7Bookmarks
Is that a lot?
0.95×vs this author's median6 550 views is typical
40Percentile for this authorof 10 recent posts
1.09×vs under 10K median5 664 views is typical
141.12%Reachviews ÷ followers
1.10%Engagement rateof viewers reacted