RunsOnMy ModelsGPUs & MacsAPI pricesVRAM calculatorNew

Can I run glm 4 9b locally?

glm 4 9b by zai-org has 9.4B parameters. At 4-bit with an 8K context it needs about 11.5 GB of GPU memory. It fits on 59 of 65 GPUs and Macs we track.

11.5 GBQ4_K_M · 8K context
RTX 4060 Ti 16GBsmallest consumer GPU that fits
M2 24GBsmallest Mac that fits
8Kmax context
MHAattention type

Memory needed by quant and context length

GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use real file sizes from mradermacher/glm-4-9b-GGUF where available (marked •), otherwise parameter count × bits per weight. How we calculate.

QuantQuality4K ctx8K ctx
2-bit (Q2_K) •rough6.9 GB9.4 GB
3-bit (Q3_K_M) •usable7.9 GB10.4 GB
4-bit (Q4_K_M) •good, most popular9.0 GB11.5 GB
5-bit (Q5_K_M) •very good9.8 GB12.3 GB
6-bit (Q6_K) •near lossless10.9 GB13.4 GB
8-bit (Q8_0) •effectively lossless12.5 GB15.0 GB
16-bit (F16/BF16) •original weights20.7 GB23.2 GB

Which GPUs and Macs run glm 4 9b?

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 GBFitsQ2_K9.4 GB90details
RTX 4060 Ti 16GB16 GBFitsQ6_K13.4 GB35details
RTX 309024 GBFitsQ8_015.0 GB94details
RTX 409024 GBFitsQ8_015.0 GB101details
RTX 509032 GBFitsF1623.2 GB95details
RTX 5070 Ti16 GBFitsQ6_K13.4 GB108details
RX 7900 XTX24 GBFitsQ8_015.0 GB96details
Mac M4 Pro 48GB48 GBFitsF1623.2 GB15details
Mac M4 Max 128GB128 GBFitsF1623.2 GB29details
Mac M3 Ultra 512GB512 GBFitsF1623.2 GB44details
DGX Spark128 GBFitsF1623.2 GB15details
RTX 407012 GBFitsQ2_K9.4 GB126details
RTX 5060 Ti 16GB16 GBFitsQ6_K13.4 GB54details
RTX 40608 GBNo———details
Arc B58012 GBFitsQ2_K9.4 GB114details
RTX 3060 Ti8 GBNo———details
RTX 30708 GBNo———details
RTX 3080 10GB10 GBTightQ2_K9.4 GB190details
RTX 3080 12GB12 GBFitsQ2_K9.4 GB228details
RTX 3090 Ti24 GBFitsQ8_015.0 GB101details
RTX 4060 Ti 8GB8 GBNo———details
RTX 4070 Super12 GBFitsQ2_K9.4 GB126details
RTX 4070 Ti12 GBFitsQ2_K9.4 GB126details
RTX 4070 Ti Super16 GBFitsQ6_K13.4 GB81details
RTX 408016 GBFitsQ6_K13.4 GB87details
RTX 4080 Super16 GBFitsQ6_K13.4 GB89details
RTX 50608 GBNo———details
RTX 507012 GBFitsQ2_K9.4 GB168details
RTX 508016 GBFitsQ6_K13.4 GB116details
2x RTX 30902×24 GBFitsF1623.2 GB50details
2x RTX 40902×24 GBFitsF1623.2 GB54details
2x RTX 50902×32 GBFitsF1623.2 GB95details
RTX A600048 GBFitsF1623.2 GB41details
RTX 6000 Ada48 GBFitsF1623.2 GB51details
RTX PRO 6000 Blackwell96 GBFitsF1623.2 GB95details
L40S48 GBFitsF1623.2 GB46details
A100 40GB40 GBFitsF1623.2 GB83details
A100 80GB80 GBFitsF1623.2 GB108details
H100 PCIe80 GBFitsF1623.2 GB106details
H100 SXM80 GBFitsF1623.2 GB178details
H200141 GBFitsF1623.2 GB255details
RX 76008 GBNo———details
RX 7700 XT12 GBFitsQ2_K9.4 GB108details
RX 7800 XT16 GBFitsQ6_K13.4 GB76details
RX 7900 GRE16 GBFitsQ6_K13.4 GB70details
RX 7900 XT20 GBFitsQ8_015.0 GB80details
RX 907016 GBFitsQ6_K13.4 GB77details
RX 9070 XT16 GBFitsQ6_K13.4 GB77details
Ryzen AI Max+ 395 128GB128 GBFitsF1623.2 GB14details
Arc A770 16GB16 GBFitsQ6_K13.4 GB68details
Mac M1 16GB16 GBFitsQ2_K9.4 GB17details
Mac M2 24GB24 GBFitsQ6_K13.4 GB12details
Mac M3 24GB24 GBFitsQ6_K13.4 GB12details
Mac M4 16GB16 GBFitsQ2_K9.4 GB30details
Mac M4 32GB32 GBFitsQ8_015.0 GB12details
Mac M1 Pro 32GB32 GBFitsQ8_015.0 GB20details
Mac M3 Pro 36GB36 GBFitsQ8_015.0 GB15details
Mac M4 Pro 24GB24 GBFitsQ6_K13.4 GB33details
Mac M4 Pro 64GB64 GBFitsF1623.2 GB15details
Mac M1 Max 64GB64 GBFitsF1623.2 GB21details
Mac M2 Max 96GB96 GBFitsF1623.2 GB21details
Mac M3 Max 128GB128 GBFitsF1623.2 GB21details
Mac M4 Max 64GB64 GBFitsF1623.2 GB29details
Mac M2 Ultra 192GB192 GBFitsF1623.2 GB43details
Mac M3 Ultra 256GB256 GBFitsF1623.2 GB44details

Questions

How much VRAM does glm 4 9b need?

About 11.5 GB at 4-bit (Q4_K_M) with an 8K context: 5.8 GB of weights (real GGUF file size), 5.0 GB of KV cache and 0.7 GB of runtime overhead. At 8-bit it needs about 15.0 GB.

What is the smallest GPU that runs glm 4 9b?

At Q4_K_M and 8K context, the smallest single consumer GPU in our table that fits it is the RTX 4060 Ti 16GB (16 GB).

Can I run glm 4 9b on a Mac?

Yes. At Q4_K_M with 8K context it fits on a Mac M2 24GB or larger, using the default macOS GPU memory limit.

Model facts

Hugging Face repozai-org/glm-4-9b
Parameters9,399,951,392 (9.4B)
Layers / hidden size40 / 4096
AttentionMHA; 32 heads, 32 KV heads, head dim 128
KV cache per token (FP16)640.0 KiB
Max context8,192 tokens
Licenseother
Released on HFJun 4, 2024
Downloads (30d) / likes22,983 / 145

Sources: Hugging Face model API · config.json (zai-org/glm-4-9b) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.