RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can I run GLM 4 32B 0414 locally?

GLM 4 32B 0414 by zai-org has 33B parameters. At 4-bit with an 8K context it needs about 19.5 GB of GPU memory. It fits on 48 of 65 GPUs and Macs we track. See all models that fit in 24 GB of VRAM.

19.5 GBQ4_K_M · 8K context
RTX 3090smallest consumer GPU that fits
M3 Pro 36GBsmallest Mac that fits
32Kmax context
GQAattention type

Memory needed by quant and context length

GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use real file sizes from axteam/GLM-4-32B-0414-GGUF where available (marked •), otherwise parameter count × bits per weight. How we calculate.

QuantQuality4K ctx8K ctx32K ctx
2-bit (Q2_K) •rough12.4 GB12.6 GB14.0 GB
3-bit (Q3_K_M) •usable15.7 GB16.0 GB17.4 GB
4-bit (Q4_K_M) •good, most popular19.3 GB19.5 GB20.9 GB
5-bit (Q5_K_M) •very good22.4 GB22.7 GB24.1 GB
6-bit (Q6_K) •near lossless25.8 GB26.1 GB27.5 GB
8-bit (Q8_0) •effectively lossless33.2 GB33.4 GB34.8 GB
16-bit (F16/BF16)original weights61.6 GB61.8 GB63.3 GB

Which GPUs and Macs run GLM 4 32B 0414?

Best quality quant that fits at 8K context. "Tight" = within 10% of usable memory. Speed is a ceiling from memory bandwidth (real speed is typically 50–80% of it).

HardwareMemoryFits?Best quantNeedsMax tok/s (est.)
RTX 3060 12GB12 GBNo———details
RTX 4060 Ti 16GB16 GBFitsQ2_K12.6 GB23details
RTX 309024 GBFitsQ4_K_M19.5 GB48details
RTX 409024 GBFitsQ4_K_M19.5 GB51details
RTX 509032 GBFitsQ6_K26.1 GB67details
RTX 5070 Ti16 GBFitsQ2_K12.6 GB73details
RX 7900 XTX24 GBFitsQ4_K_M19.5 GB49details
Mac M4 Pro 48GB48 GBFitsQ6_K26.1 GB10details
Mac M4 Max 128GB128 GBFitsF1661.8 GB8.4details
Mac M3 Ultra 512GB512 GBFitsF1661.8 GB13details
DGX Spark128 GBFitsF1661.8 GB4.2details
RTX 407012 GBNo———details
RTX 5060 Ti 16GB16 GBFitsQ2_K12.6 GB36details
RTX 40608 GBNo———details
Arc B58012 GBNo———details
RTX 3060 Ti8 GBNo———details
RTX 30708 GBNo———details
RTX 3080 10GB10 GBNo———details
RTX 3080 12GB12 GBNo———details
RTX 3090 Ti24 GBFitsQ4_K_M19.5 GB51details
RTX 4060 Ti 8GB8 GBNo———details
RTX 4070 Super12 GBNo———details
RTX 4070 Ti12 GBNo———details
RTX 4070 Ti Super16 GBFitsQ2_K12.6 GB55details
RTX 408016 GBFitsQ2_K12.6 GB58details
RTX 4080 Super16 GBFitsQ2_K12.6 GB60details
RTX 50608 GBNo———details
RTX 507012 GBNo———details
RTX 508016 GBFitsQ2_K12.6 GB78details
2x RTX 30902×24 GBFitsQ8_033.4 GB27details
2x RTX 40902×24 GBFitsQ8_033.4 GB29details
2x RTX 50902×32 GBFitsQ8_033.4 GB52details
RTX A600048 GBFitsQ8_033.4 GB22details
RTX 6000 Ada48 GBFitsQ8_033.4 GB28details
RTX PRO 6000 Blackwell96 GBFitsF1661.8 GB28details
L40S48 GBFitsQ8_033.4 GB25details
A100 40GB40 GBFitsQ8_033.4 GB45details
A100 80GB80 GBFitsF1661.8 GB31details
H100 PCIe80 GBFitsF1661.8 GB31details
H100 SXM80 GBFitsF1661.8 GB51details
H200141 GBFitsF1661.8 GB74details
RX 76008 GBNo———details
RX 7700 XT12 GBNo———details
RX 7800 XT16 GBFitsQ2_K12.6 GB51details
RX 7900 GRE16 GBFitsQ2_K12.6 GB47details
RX 7900 XT20 GBFitsQ3_K_M16.0 GB50details
RX 907016 GBFitsQ2_K12.6 GB52details
RX 9070 XT16 GBFitsQ2_K12.6 GB52details
Ryzen AI Max+ 395 128GB128 GBFitsF1661.8 GB3.9details
Arc A770 16GB16 GBFitsQ2_K12.6 GB46details
Mac M1 16GB16 GBNo———details
Mac M2 24GB24 GBFitsQ2_K12.6 GB8.1details
Mac M3 24GB24 GBFitsQ2_K12.6 GB8.1details
Mac M4 16GB16 GBNo———details
Mac M4 32GB32 GBFitsQ3_K_M16.0 GB7.6details
Mac M1 Pro 32GB32 GBFitsQ3_K_M16.0 GB13details
Mac M3 Pro 36GB36 GBFitsQ4_K_M19.5 GB7.6details
Mac M4 Pro 24GB24 GBFitsQ2_K12.6 GB22details
Mac M4 Pro 64GB64 GBFitsQ8_033.4 GB7.9details
Mac M1 Max 64GB64 GBFitsQ8_033.4 GB12details
Mac M2 Max 96GB96 GBFitsF1661.8 GB6.1details
Mac M3 Max 128GB128 GBFitsF1661.8 GB6.1details
Mac M4 Max 64GB64 GBFitsQ8_033.4 GB16details
Mac M2 Ultra 192GB192 GBFitsF1661.8 GB12details
Mac M3 Ultra 256GB256 GBFitsF1661.8 GB13details

Questions

How much VRAM does GLM 4 32B 0414 need?

About 19.5 GB at 4-bit (Q4_K_M) with an 8K context: 18.3 GB of weights (real GGUF file size), 0.5 GB of KV cache and 0.7 GB of runtime overhead. At 8-bit it needs about 33.4 GB.

What is the smallest GPU that runs GLM 4 32B 0414?

At Q4_K_M and 8K context, the smallest single consumer GPU in our table that fits it is the RTX 3090 (24 GB).

Can I run GLM 4 32B 0414 on a Mac?

Yes. At Q4_K_M with 8K context it fits on a Mac M3 Pro 36GB or larger, using the default macOS GPU memory limit.

Similar models

Same family or a similar parameter count, with the VRAM each needs at Q4_K_M and 8K context.

GLM Z1 32B 0414

33B · 19.5 GB at 4-bit

GLM 4.7 Flash

31B · 18.2 GB at 4-bit

GLM 4.6V Flash

10B · 6.7 GB at 4-bit

Qwen3 32B

33B · 21.1 GB at 4-bit

DeepSeek R1 Distill Qwen 32B

33B · 21.2 GB at 4-bit

granite 4.0 h small

32B · 18.9 GB at 4-bit

Model facts

Hugging Face repozai-org/GLM-4-32B-0414
Parameters32,566,081,536 (33B)
Layers / hidden size61 / 6144
AttentionGQA; 48 heads, 2 KV heads, head dim 128
KV cache per token (FP16)61.0 KiB
Max context32,768 tokens
Licensemit
Released on HFApr 7, 2025
Downloads (30d) / likes8,324 / 491

Sources: Hugging Face model API · config.json (zai-org/GLM-4-32B-0414) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.