Gemini 3.8 Flash vs GLM 5.3 Flash
2 posts10 likes1.7K views
Gemini 3.8 Flash vs GLM 5.3 Flash in real side-by-side tests: 2 posts from X where both got the same prompt; 1 post reports cost or time. Judge coding ability from the actual outputs.
Posts
2Same prompt, separate runs
1Not a controlled comparison: prompts, tools and time differ.
Summary
Gemini 3.8 Flash and GLM 5.3 Flash appear together in 2 posts on X: tree house, two Flash models (@Rohit_115) and 3D voxel pagoda (@Rohit_115). 1 post reports numbers for only one side, so there is no like-for-like cost or time comparison yet. According to @Rohit_115, GLM 5.3 Flash nailed it; Gemini 3.8 Flash did a decent job. (The author’s view.)
What each post reports
| Post | Model | Time | Cost | Tokens | Feedback rounds | Author’s verdict | Original |
|---|---|---|---|---|---|---|---|
| Tree house, two Flash models@Rohit_115 | GLM 5.3 Flash | — | — | — | — | — | X post |
| Tree house, two Flash models@Rohit_115 | Gemini 3.8 Flash | — | — | — | — | — | X post |
| 3D voxel pagoda@Rohit_115 | GLM 5.3 Flash | — | — | — | 1 | Author’s pick | X post |
| 3D voxel pagoda@Rohit_115 | Gemini 3.8 Flash | — | — | — | 1 | GLM 5.3 Flash nailed it; Gemini 3.8 Flash did a decent job | X post |
Only numbers stated in each post. “—” means the post doesn’t say.
Who won where
1- 3D voxel pagodaGLM 5.3 Flash nailed it; Gemini 3.8 Flash did a decent job.— the author’s view, @Rohit_115 (paraphrased)