GLM-5.2 looks stronger, but token cost is still a problem
GLM-5.2’s published numbers appear close to GPT-5.5 across several tests and not far behind Opus 4.8 in many areas. That would be a notable jump for a Chinese model compared with roughly six months ago. The weaker point is in tasks.
GLM-5.2 Max is shown using almost twice as many tokens as Opus 4.7 Max for a similar score. still looks much better when score is compared with token use. In practical use, cheaper Chinese models can take over some work to reduce API spending, but Claude is still preferred for harder and prompts with many conditions.
GLM-5.2 may also be slower on larger jobs and use more tokens, so it is better seen as a way to split work across models, not as a full replacement.
Key points
- GLM-5.2 appears close to GPT-5.5 on several results.
- Its scores look better, but its still trails s.
- GLM-5.2 Max is shown using almost twice the tokens of Opus 4.7 Max for a similar score.
- Claude remains stronger for difficult and prompts with many conditions.
- The best cost-saving use may be routing easier tasks to cheaper models, not replacing the main model entirely.