Short answer: Qwen 3.8 Flash is 2.5x cheaper (0.034¢ vs 0.086¢ a chat). GLM 5.3 FlashX ranks higher overall (#19 vs #20). Pick Qwen 3.8 Flash for everyday use and GLM 5.3 FlashX when quality matters most.


| Qwen 3.8 Flash | GLM 5.3 FlashX | |
|---|---|---|
| Cost per chat (NZD) | 0.034¢ | 0.086¢ |
| Chats per NZ$1 | 2,913 | 1,157 |
| Overall rank | #20 | #19 |
| Vendor | Alibaba | Z.ai |
| Speciality | General | Reasoning |
| Crowd Elo | no votes | 1012 |
| Reasoning | 56 | 56 |
| Code | 50 | 50 |
| Math | 52 | 52 |
| Writing | 54 | 54 |
| Instruction | 54 | 54 |
Scores are 0 to 100 percentiles blending LMArena, Epoch AI, Artificial Analysis and CrapGPT's own tests. Updated 2026-10-11.
Qwen 3.8 Flash costs 0.034¢ per typical chat versus 0.086¢ for GLM 5.3 FlashX on CrapGPT (NZD, measured token use).
GLM 5.3 FlashX ranks higher overall on CrapGPT's blended benchmark score.
Yes. CrapGPT's card battle sends one prompt to two models and lets you pick the winner, paying only for the two replies.
Battle them yourself Qwen 3.8 Flash details · GLM 5.3 FlashX details