MODEL · GPU GUIDE

GLM-5.2-NVFP4 GPU Requirements: VRAM & Cheapest GPU

GLM-5.2-NVFP4 has about 381B parameters. See exactly how much GPU memory it needs at FP16, INT8, and INT4, and the cheapest GPU to run it, with live hourly pricing from 5+ data center partners.

381BParameters
208 GBMin VRAM
$3.20/hrCheapest
< 2 minDeploy
nvidia/GLM-5.2-NVFP4
VIEW ON HUGGINGFACE ↗
381B paramstext-generationglm_moe_dsa441 downloads80 likesupdated Jun 26, 2026

To run GLM-5.2-NVFP4 for inference at FP16, you need roughly 831 GB of VRAM. The cheapest fit on Spheron is 4x B300 288GB at about $14.00/hr. Quantize to INT4 to run it on a smaller, cheaper GPU.

GB VRAM REQUIRED
FP16INFERENCEBATCH 1CTX 4k

Estimated peak VRAM including weights, activations, and KV cache. Add 10% headroom for production traffic.

RANKCONFIGURATIONPER GPUTOTAL $/HR
  • 01
    4× B300 288GBCHEAPEST
    Blackwell Ultra · HBM3e
    $3.50/hr$14.00/hr
  • 02
    8× B200 192GB
    Blackwell · HBM3e
    $2.74/hr$21.92/hr
  • 03
    8× H200 141GB
    Hopper · HBM3e
    $3.31/hr$26.48/hr

Live pricing aggregated from 5+ data center partners. Per-minute billing, no commitments.

VRAM required to run GLM-5.2-NVFP4

Estimated peak VRAM at context length 4,096 and batch size 1, including weights, activations, and KV cache. Quantizing to INT8 (Q8) or INT4 (Q4) cuts memory roughly in half and in quarter.

PrecisionInferenceLoRA fine-tuneFull fine-tune
FP16831 GB1246 GB3322 GB
INT8415 GB623 GB1661 GB
INT4208 GB311 GB831 GB

Cheapest GPU to run GLM-5.2-NVFP4 by precision

FP16
VRAM required831GB

Full precision. Best quality, highest memory.

Cheapest GPU
4x B300 288GB
Blackwell Ultra · HBM3e
$14.00/hr · $3.50/hr/gpu
4x B300 288GB on Spheron
INT8
VRAM required415GB

8-bit quantized. ~2x smaller, minimal quality loss.

Cheapest GPU
8x A100 80GB
Ampere · HBM2e
$6.40/hr · $0.80/hr/gpu
8x A100 80GB on Spheron
INT4
VRAM required208GB

4-bit quantized. ~4x smaller, runs on smaller GPUs.

Cheapest GPU
4x A100 80GB
Ampere · HBM2e
$3.20/hr · $0.80/hr/gpu
4x A100 80GB on Spheron

Inference vs fine-tuning GLM-5.2-NVFP4

InferenceWeights + KV cache
LoRA fine-tune~1.5×+ low-rank adapter
Full fine-tune~4×+ gradients + optimizer state

Inference only holds the model weights plus a KV cache, so it is the cheapest setup. LoRA fine-tuning adds a small adapter and roughly 50% more memory. Full fine-tuning holds gradients and optimizer state on top of the weights, which is about 4x the inference footprint, so it often needs multiple GPUs even when inference fits on one. For GLM-5.2-NVFP4, an on-demand B300 288GB instance covers inference and LoRA, while a full fine-tune needs several times that memory and often spans multiple GPUs. Check the live GPU pricing for current rates.

Similar models

Compare GPU requirements for models in the same class.

FAQ / 05

GLM-5.2-NVFP4 GPU questions