GLM-5.2
#49 of 274 scored.
Z.aiJun 16, 2026Open weights
Capabilities Index
Range 149.8–153.8
151.8
Context
Takes text
1.05M
Input
Per 1M tokens
$0.06
Output
Per 1M tokens
$7.00
Benchmarks
How it scores.
- GPQA Diamond89.1%
24th of 186
- SWE-bench Verified78.7%
5th of 32
- FrontierMath59.2%
35th of 81
- ARC-AGI-222.8%
37th of 79
- SimpleQA Verified34.2%
52nd of 77
- Mock AIME86.4%
67th of 176
Everything else Epoch has for it
- DTBench89.3%
- ARC-AGI77.0%
- WeirdML70.1%
- Surface Evolver Bench55.6%
- LMCA53.9%
- SimpleBench50.6%
- DeepSWE43.8%
- ProofBench35.0%
- PostTrainBench31.7%
- FrontierMath-Tier-4-v2-Private29.3%
- FrontierCode24.5%
- Chess Puzzles16.9%
- Mystery Game Puzzles10.8%
- EBR-bench9.5%
On the index
Around it.
Sources