HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

GLM-5.2-FP8

zai-org/GLM-5.2-FP8

kind
fp8
quant
fp8 (8-bit)
family
glm
modality
text-generation
root model
zai-org/GLM-5.2
provider
zai-org
license
mit
language
en, zh
context
300,000
library
transformers
vRAM (GB)
—
size on disk
—
downloads
3,142,483
likes
239
created
—
last modified
2026-07-02T08:09:27+00:00 (2026)
huggingface
https://huggingface.co/zai-org/GLM-5.2-FP8
tags
transformerssafetensorsglm_moe_dsatext-generationconversationalenzharxiv:2602.15763arxiv:2603.12201license:mitendpoints_compatiblefp8region:us

Benchmark scores (9)

benchmarkcategoryscorecontextconfevidence
AIME 2026math99.2AIME 2026medium| AIME 2026 | 99.2 | 95.3 | 97 | - | 94.6 | 95.7 | 98.3 | 98.2 |
GPQA Diamondreasoning91.2GPQA-Diamondmedium| GPQA-Diamond | 91.2 | 86.2 | 90 | 93 | 90.1 | 93.6 | 93.6 | 94.3 |
Terminal-Bench 2.1agentic-coding82.7Terminal Bench 2.1 (Best Reported Harness)medium| Terminal Bench 2.1 (Best Reported Harness) | 82.7 | 69 | - | - | - | 78.9 | 83.4 | 70.7 |
Terminal-Bench 2.1agentic-coding81.0Terminal Bench 2.1 (Terminus-2)medium| Terminal Bench 2.1 (Terminus-2) | 81.0 | 63.5 | 75 | 65 | 64 | 85 | 84 | 74 |
FrontierSWEswe74.4FrontierSWE (Dominance)medium| FrontierSWE (Dominance) | 74.4 | 30.5 | - | - | 29.0 | 75.1 | 72.6 | 39.6 |
Humanity's Last Examknowledge54.7HLE (w/ Tools)medium| HLE (w/ Tools) | 54.7 | 52.3 | 53.5 | - | 48.2 | 57.9* | 52.2* | 51.4* |
DeepSWEswe46.2DeepSWEmedium| DeepSWE | 46.2 | 18 | 18 | 20 | 8 | 58 | 70 | 10 |
Humanity's Last Examknowledge40.5HLEmedium| HLE | 40.5 | 31 | 41.4 | 37 | 37.7 | 49.8* | 41.4* | 45 |
Terminal-Benchagentic-coding2.1mediumTerminal Bench 2.1

Discussions (1)

titleauthorstatuscommentstype
glm5.2Vaishradraft0disc
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.