RunsOnMy ModelsGPUs & MacsAPI pricesVRAM calculatorNew

Can I run LFM3 1.7B Base locally?

LFM3 1.7B Base by LiquidAI has 1.7B parameters. At 4-bit with an 8K context it needs about 1.7 GB of GPU memory. It fits on 65 of 65 GPUs and Macs we track.

1.7 GBQ4_K_M · 8K context
RTX 4060smallest consumer GPU that fits
M1 16GBsmallest Mac that fits
256Kmax context
GQA + sliding windowattention type

GGUF Q8_0 file was 395% of the size expected from the parameter count, so this page uses size estimates instead of real file sizes.

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 ctx128K ctx
2-bit (Q2_K)rough1.3 GB1.3 GB1.6 GB2.8 GB
3-bit (Q3_K_M)usable1.5 GB1.5 GB1.8 GB2.9 GB
4-bit (Q4_K_M)good, most popular1.7 GB1.7 GB2.0 GB3.1 GB
5-bit (Q5_K_M)very good1.8 GB1.9 GB2.2 GB3.3 GB
6-bit (Q6_K)near lossless2.0 GB2.1 GB2.3 GB3.5 GB
8-bit (Q8_0)effectively lossless2.4 GB2.5 GB2.7 GB3.9 GB
16-bit (F16/BF16)original weights3.9 GB4.0 GB4.2 GB5.4 GB

Which GPUs and Macs run LFM3 1.7B Base?

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 GBFitsF164.0 GB104details
RTX 4060 Ti 16GB16 GBFitsF164.0 GB83details
RTX 309024 GBFitsF164.0 GB271details
RTX 409024 GBFitsF164.0 GB292details
RTX 509032 GBFitsF164.0 GB518details
RTX 5070 Ti16 GBFitsF164.0 GB259details
RX 7900 XTX24 GBFitsF164.0 GB278details
Mac M4 Pro 48GB48 GBFitsF164.0 GB79details
Mac M4 Max 128GB128 GBFitsF164.0 GB158details
Mac M3 Ultra 512GB512 GBFitsF164.0 GB237details
DGX Spark128 GBFitsF164.0 GB79details
RTX 407012 GBFitsF164.0 GB146details
RTX 5060 Ti 16GB16 GBFitsF164.0 GB130details
RTX 40608 GBFitsF164.0 GB79details
Arc B58012 GBFitsF164.0 GB132details
RTX 3060 Ti8 GBFitsF164.0 GB130details
RTX 30708 GBFitsF164.0 GB130details
RTX 3080 10GB10 GBFitsF164.0 GB220details
RTX 3080 12GB12 GBFitsF164.0 GB264details
RTX 3090 Ti24 GBFitsF164.0 GB292details
RTX 4060 Ti 8GB8 GBFitsF164.0 GB83details
RTX 4070 Super12 GBFitsF164.0 GB146details
RTX 4070 Ti12 GBFitsF164.0 GB146details
RTX 4070 Ti Super16 GBFitsF164.0 GB194details
RTX 408016 GBFitsF164.0 GB207details
RTX 4080 Super16 GBFitsF164.0 GB213details
RTX 50608 GBFitsF164.0 GB130details
RTX 507012 GBFitsF164.0 GB194details
RTX 508016 GBFitsF164.0 GB278details
2x RTX 30902×24 GBFitsF164.0 GB271details
2x RTX 40902×24 GBFitsF164.0 GB292details
2x RTX 50902×32 GBFitsF164.0 GB518details
RTX A600048 GBFitsF164.0 GB222details
RTX 6000 Ada48 GBFitsF164.0 GB278details
RTX PRO 6000 Blackwell96 GBFitsF164.0 GB518details
L40S48 GBFitsF164.0 GB250details
A100 40GB40 GBFitsF164.0 GB450details
A100 80GB80 GBFitsF164.0 GB590details
H100 PCIe80 GBFitsF164.0 GB578details
H100 SXM80 GBFitsF164.0 GB969details
H200141 GBFitsF164.0 GB1388details
RX 76008 GBFitsF164.0 GB83details
RX 7700 XT12 GBFitsF164.0 GB125details
RX 7800 XT16 GBFitsF164.0 GB180details
RX 7900 GRE16 GBFitsF164.0 GB167details
RX 7900 XT20 GBFitsF164.0 GB231details
RX 907016 GBFitsF164.0 GB185details
RX 9070 XT16 GBFitsF164.0 GB185details
Ryzen AI Max+ 395 128GB128 GBFitsF164.0 GB74details
Arc A770 16GB16 GBFitsF164.0 GB162details
Mac M1 16GB16 GBFitsF164.0 GB20details
Mac M2 24GB24 GBFitsF164.0 GB29details
Mac M3 24GB24 GBFitsF164.0 GB29details
Mac M4 16GB16 GBFitsF164.0 GB35details
Mac M4 32GB32 GBFitsF164.0 GB35details
Mac M1 Pro 32GB32 GBFitsF164.0 GB58details
Mac M3 Pro 36GB36 GBFitsF164.0 GB43details
Mac M4 Pro 24GB24 GBFitsF164.0 GB79details
Mac M4 Pro 64GB64 GBFitsF164.0 GB79details
Mac M1 Max 64GB64 GBFitsF164.0 GB116details
Mac M2 Max 96GB96 GBFitsF164.0 GB116details
Mac M3 Max 128GB128 GBFitsF164.0 GB116details
Mac M4 Max 64GB64 GBFitsF164.0 GB158details
Mac M2 Ultra 192GB192 GBFitsF164.0 GB231details
Mac M3 Ultra 256GB256 GBFitsF164.0 GB237details

Questions

How much VRAM does LFM3 1.7B Base need?

About 1.7 GB at 4-bit (Q4_K_M) with an 8K context: 1.0 GB of weights (estimated), 0.1 GB of KV cache and 0.6 GB of runtime overhead. At 8-bit it needs about 2.5 GB.

What is the smallest GPU that runs LFM3 1.7B Base?

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

Can I run LFM3 1.7B Base on a Mac?

Yes. At Q4_K_M with 8K context it fits on a Mac M1 16GB or larger, using the default macOS GPU memory limit.

Model facts

Hugging Face repoLiquidAI/LFM3-1.7B-Base
Parameters1,728,801,536 (1.7B)
Layers / hidden size24 / 2048
AttentionGQA + sliding window (hybrid: 13 linear/state layers); 16 heads, 4 KV heads, head dim 128; sliding window 256
KV cache per token (FP16)22.0 KiB (sliding layers capped at the window)
Max context262,144 tokens
Licenseother
Released on HFOct 26, 2026
Downloads (30d) / likes12,253 / 164

Sources: Hugging Face model API · config.json (LiquidAI/LFM3-1.7B-Base) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.