Short answer: Qwen 3.8 Max is 9.2x cheaper (0.36¢ vs 3.35¢ a chat). Claude Opus 5 ranks higher overall (#2 vs #3). Qwen 3.8 Max wins at agentic tasks. Claude Opus 5 wins at reasoning, coding, maths, writing, following instructions. Pick Qwen 3.8 Max for everyday use and Claude Opus 5 when quality matters most.


| Qwen 3.8 Max | Claude Opus 5 | |
|---|---|---|
| Cost per chat (NZD) | 0.36¢ | 3.35¢ |
| Chats per NZ$1 | 275 | 29 |
| Overall rank | #3 | #2 |
| Vendor | Alibaba | Anthropic |
| Speciality | General | General |
| Crowd Elo | 988 | no votes |
| Reasoning | 89 | 90 |
| Code | 82 | 88 |
| Math | 83 | 86 |
| Writing | 84 | 93 |
| Instruction | 80 | 88 |
| Agentic | 97 | 92 |
Scores are 0 to 100 percentiles blending LMArena, Epoch AI, Artificial Analysis and CrapGPT's own tests. Updated 2026-10-11.
Qwen 3.8 Max costs 0.36¢ per typical chat versus 3.35¢ for Claude Opus 5 on CrapGPT (NZD, measured token use).
Claude Opus 5 ranks higher overall on CrapGPT's blended benchmark score. By category: Qwen 3.8 Max leads at agentic tasks; Claude Opus 5 leads at reasoning, coding, maths, writing, following instructions.
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 Max details · Claude Opus 5 details