HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/zai-org/zai-org/GLM-5.2

zai-org/GLM-5.2

family glm · 6 variants · unknown

provider
zai-org
released
—
huggingface
https://huggingface.co/zai-org/GLM-5.2
best scores
Terminal-Bench: 2.1Humanity's Last Exam: 54.7AIME 2026: 99.2GPQA Diamond: 91.2DeepSWE: 46.2Terminal-Bench 2.1: 82.7FrontierSWE: 74.4

Best scores per benchmark

benchmarkcategorybest score
AIME 2026math99.2
GPQA Diamondreasoning91.2
Terminal-Bench 2.1agentic-coding82.7
FrontierSWEswe74.4
Humanity's Last Examknowledge54.7
DeepSWEswe46.2
Terminal-Benchagentic-coding2.1

Variants (6)

variantkindquantbitsvRAMdlorgyear
GLM-5.2-FP8fp8fp88.03,142,483zai-org2026
GLM-5.2-NVFP4nvfp4nvfp44.01,607,192nvidia2026
GLM-5.2originalbf1616.01,003,547zai-org2026
GLM-5.2-GGUFggufgguf16.0363,121unsloth2026
GLM-5.2-FP8fp8fp88.01,031unsloth2026
GLM-5.2originalbf1616.0467unsloth2026
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.