Bijan Bowen Tests GLM 5.2 for Coding and 3D

Bijan Bowen33m
Bijan Bowen beside the blue and white headline GLM 5.2 Tested.
0 comments · 0 votesOpen discussionClose discussion
Sign in to join the discussion

    Bijan Bowen tests GLM 5.2 while it is initially limited to coding-plan users. He describes an announced future API and open-weight release as plans at the time of recording.

    His browser OSCode generation uses AI or another automated system to create source code from instructions, examples, schemas, or higher-level specifications. includes a playable city scene and a striking 3D window effect, although other components fail. He also tests a skating game without Raylib and praises a separate rally game that uses it, while noting visual defects.

    Further tasks include a subway game, a watch website, a printer simulation and a printable V8 model. Iteration improves some results, but the printer simulation still fails; his overall ranking of the model is an opinion based on these informal testsEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria..

    Original YouTube thumbnailWatch on YouTube

    Video topic coverage

    • Agents & workflows30
    • Coding & tools100
    • Models & learning90
    • Infrastructure & compute20
    • Research & evaluation100
    • Safety & security0
    • Business & productivity20
    • Creative & media0
    • Robotics & perception0
    • Society & policy0

    Share this video

    Bijan Bowen beside the blue and white headline GLM 5.2 Tested. With WWW.ARTIFICIAL-INTELLIGENCE.VIDEO, 13 June 2026 and duration 33m.

    Bijan Bowen tries GLM 5.2 on varied coding and 3D tasks, finding strong results alongside unfinished and failed features.