Skip to main contentThe Local Bench

Tool use and agents · Z.ai

GLM-4.5 Air

Hugging Face: zai-org/GLM-4.5-Air · checked 2026-10-13

106B MoE (12B active) agent model; needs 64 GB+ (unified memory or multi-GPU).

Datasheet

Maker
Z.ai
Parameters
110.47 BVerifiedsafetensors.total from the HF API Source
Licence
MITVerified
Access
Public download
Task (HF pipeline)
text-generation
First published on HF
2025-07-20
Last updated on HF
2025-08-11
Weights in memory
221 GB at 16-bit · 67 GB at ~4.85 bitsEstimateParameters × bits ÷ 8. Excludes context cache and runtime overhead; not every model has a 4-bit build.

Benchmark

Not yet measured
BFCL V4, overall accuracy

No comparable public score for this benchmark yet (Not listed in BFCL V4). We will not estimate one.

Other published results

  • LMArena text, coding category: 1,397.1Sourced Source

The description summarises the model card at https://huggingface.co/zai-org/GLM-4.5-Air. Parameters and licence are read from the Hugging Face API; the weights-in-memory line is our arithmetic, not a measurement.