RunsOnMy ModelsGPUs & MacsAPI pricesVRAM calculatorNew

Can I run LLaDA2.1 mini locally?

LLaDA2.1 mini by inclusionAI has 16B parameters (1.4B active per token, 256 experts, 8 used per token). At 4-bit with an 8K context it needs about 10.3 GB of GPU memory. It fits on 65 of 65 GPUs and Macs we track.

10.3 GBQ4_K_M · 8K context
RTX 4060 Ti 16GBsmallest consumer GPU that fits
M2 24GBsmallest 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 parameter count × bits per weight. How we calculate.

QuantQuality4K ctx8K ctx32K ctx
2-bit (Q2_K)rough6.5 GB6.7 GB7.6 GB
3-bit (Q3_K_M)usable8.3 GB8.4 GB9.3 GB
4-bit (Q4_K_M)good, most popular10.1 GB10.3 GB11.2 GB
5-bit (Q5_K_M)very good11.6 GB11.8 GB12.7 GB
6-bit (Q6_K)near lossless13.3 GB13.4 GB14.4 GB
8-bit (Q8_0)effectively lossless16.9 GB17.1 GB18.0 GB
16-bit (F16/BF16)original weights31.1 GB31.3 GB32.2 GB

Which GPUs and Macs run LLaDA2.1 mini?

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 GBFitsQ3_K_M8.4 GB514details
RTX 4060 Ti 16GB16 GBFitsQ6_K13.4 GB245details
RTX 309024 GBFitsQ8_017.1 GB615details
RTX 409024 GBFitsQ8_017.1 GB662details
RTX 509032 GBFitsQ8_017.1 GB1177details
RTX 5070 Ti16 GBFitsQ6_K13.4 GB763details
RX 7900 XTX24 GBFitsQ8_017.1 GB631details
Mac M4 Pro 48GB48 GBFitsQ8_017.1 GB179details
Mac M4 Max 128GB128 GBFitsF1631.3 GB191details
Mac M3 Ultra 512GB512 GBFitsF1631.3 GB286details
DGX Spark128 GBFitsF1631.3 GB95details
RTX 407012 GBFitsQ3_K_M8.4 GB720details
RTX 5060 Ti 16GB16 GBFitsQ6_K13.4 GB381details
RTX 40608 GBFitsQ2_K6.7 GB506details
Arc B58012 GBFitsQ3_K_M8.4 GB651details
RTX 3060 Ti8 GBFitsQ2_K6.7 GB834details
RTX 30708 GBFitsQ2_K6.7 GB834details
RTX 3080 10GB10 GBFitsQ3_K_M8.4 GB1085details
RTX 3080 12GB12 GBFitsQ3_K_M8.4 GB1302details
RTX 3090 Ti24 GBFitsQ8_017.1 GB662details
RTX 4060 Ti 8GB8 GBFitsQ2_K6.7 GB536details
RTX 4070 Super12 GBFitsQ3_K_M8.4 GB720details
RTX 4070 Ti12 GBFitsQ3_K_M8.4 GB720details
RTX 4070 Ti Super16 GBFitsQ6_K13.4 GB572details
RTX 408016 GBFitsQ6_K13.4 GB610details
RTX 4080 Super16 GBFitsQ6_K13.4 GB626details
RTX 50608 GBFitsQ2_K6.7 GB834details
RTX 507012 GBFitsQ3_K_M8.4 GB959details
RTX 508016 GBFitsQ6_K13.4 GB817details
2x RTX 30902×24 GBFitsF1631.3 GB327details
2x RTX 40902×24 GBFitsF1631.3 GB352details
2x RTX 50902×32 GBFitsF1631.3 GB625details
RTX A600048 GBFitsF1631.3 GB268details
RTX 6000 Ada48 GBFitsF1631.3 GB335details
RTX PRO 6000 Blackwell96 GBFitsF1631.3 GB625details
L40S48 GBFitsF1631.3 GB301details
A100 40GB40 GBFitsF1631.3 GB543details
A100 80GB80 GBFitsF1631.3 GB711details
H100 PCIe80 GBFitsF1631.3 GB698details
H100 SXM80 GBFitsF1631.3 GB1169details
H200141 GBFitsF1631.3 GB1675details
RX 76008 GBFitsQ2_K6.7 GB536details
RX 7700 XT12 GBFitsQ3_K_M8.4 GB617details
RX 7800 XT16 GBFitsQ6_K13.4 GB531details
RX 7900 GRE16 GBFitsQ6_K13.4 GB490details
RX 7900 XT20 GBFitsQ8_017.1 GB525details
RX 907016 GBFitsQ6_K13.4 GB545details
RX 9070 XT16 GBFitsQ6_K13.4 GB545details
Ryzen AI Max+ 395 128GB128 GBFitsF1631.3 GB89details
Arc A770 16GB16 GBFitsQ6_K13.4 GB477details
Mac M1 16GB16 GBFitsQ3_K_M8.4 GB97details
Mac M2 24GB24 GBFitsQ6_K13.4 GB85details
Mac M3 24GB24 GBFitsQ6_K13.4 GB85details
Mac M4 16GB16 GBFitsQ3_K_M8.4 GB171details
Mac M4 32GB32 GBFitsQ8_017.1 GB79details
Mac M1 Pro 32GB32 GBFitsQ8_017.1 GB131details
Mac M3 Pro 36GB36 GBFitsQ8_017.1 GB99details
Mac M4 Pro 24GB24 GBFitsQ6_K13.4 GB232details
Mac M4 Pro 64GB64 GBFitsF1631.3 GB95details
Mac M1 Max 64GB64 GBFitsF1631.3 GB140details
Mac M2 Max 96GB96 GBFitsF1631.3 GB140details
Mac M3 Max 128GB128 GBFitsF1631.3 GB140details
Mac M4 Max 64GB64 GBFitsF1631.3 GB191details
Mac M2 Ultra 192GB192 GBFitsF1631.3 GB279details
Mac M3 Ultra 256GB256 GBFitsF1631.3 GB286details

Questions

How much VRAM does LLaDA2.1 mini need?

About 10.3 GB at 4-bit (Q4_K_M) with an 8K context: 9.3 GB of weights (estimated), 0.3 GB of KV cache and 0.7 GB of runtime overhead. At 8-bit it needs about 17.1 GB.

What is the smallest GPU that runs LLaDA2.1 mini?

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 LLaDA2.1 mini 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 repoinclusionAI/LLaDA2.1-mini
Parameters16,255,643,392 (16B) · ~1.4B active (computed from config)
Layers / hidden size20 / 2048
AttentionGQA; 16 heads, 4 KV heads, head dim 128
KV cache per token (FP16)40.0 KiB
Max context32,768 tokens
Licenseapache-2.0
Released on HFFeb 9, 2026
Downloads (30d) / likes27,642 / 128

Sources: Hugging Face model API · config.json (inclusionAI/LLaDA2.1-mini) · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.