All benchmarks
Zhipu / arena-v1
GLM-5.2 benchmark.
Blind preference, cited external scores, and evaluation economics. Every result keeps its sample count and source date attached.
Evidence levellow
Last updatePending
Categories1
Combined quality
Pending
Arena score
Pending
Median latency
Pending
Valid votes
0
Task analysis
Performance by task.
| Category | Quality | Arena | External | Evidence |
|---|---|---|---|---|
| overall | Pending | Pending | Pending | low / 0 votes |
Provenance
External evidence.
No external snapshot is published for this model. Its current signal comes from RouterPlex Arena votes.
Your workload is the final benchmark
Test GLM-5.2 yourself.
Use it directly, or put it into a blind head-to-head comparison.