ModelsCompareBest forBenchmarksStatusPricingAPI

GLM-5.2

Z.ai·Pending·Released 2026-06-24text
Context1,000k
Max output131.072k
Input / 1M$1.4 ↗
Output / 1M$4.4 ↗
Cached in$0.26
Benchmarks26

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

AIME 2026 (MathArena)#21/2190% ↗
FrontierMath Tier 4 v2 (Epoch AI run)#19/4429.3% ↗
FrontierMath Tiers 1-3 v2 (Epoch AI run)#25/5959.2% ↗
GPQA Diamond (Epoch AI run)#16/7791.9% ↗
OTIS Mock AIME 2024-2025 (Epoch AI run)#44/7686.4% ↗
SimpleQA Verified (Epoch AI run)#34/5334.2% ↗
SWE-bench Verified (Epoch AI run)#3/2478.7% ↗
Agents' Last Exam (ALE-CLI)#3/323.8% ↗
AutomationBench v1.0.6#3/326.2% ↗
CyberGym#2/277.2% ↗
DeepSWE v1.1#13/1346.2% ↗
ExploitBench#2/224.4% ↗
ExploitGym 2h#2/229 ↗
ExploitGym 6h#2/239 ↗
FrontierSWE#3/367.5% ↗
GDPval-AA v2#6/71508 ↗
HLE (with tools)#3/454.7% ↗
NL2Repo#3/448.9% ↗
PostTrainBench#3/331.7% ↗
ProgramBench (Almost Solved)#2/29.5% ↗
SWE-Marathon v1.1#2/219.4% ↗
Terminal-Bench 2.1#7/1781% ↗
Terminal-Bench 3.0#3/34.6% ↗
Toolathlon Verified#4/459.9% ↗
SWE-Bench Pro#7/1562.1% ↗
Terminal-Bench 2.1#7/1781% ↗

Compare

vs GPT-6 Astravs GPT-5.6 Solvs GPT-5.6 Terravs GPT-5.6 Lunavs Claude Fable 5.1vs Claude Opus 5vs Claude Sonnet 5vs Claude Haiku 4.5

More from Z.ai

GLM-5.3GLM-5.3-FlashGLM-5GLM-4.7-FlashXGLM-5.1GLM-5-Turbo

Access GLM-5.2 and every other model through one endpoint with automatic failover: Respan gateway.