LFM2-8B-A1B GGUF size and VRAM requirements
unsloth/LFM2-8B-A1B-GGUF is a large language model with 8.34 billion parameters, built on the lfm2moe architecture. It is released under the other license and has been downloaded 11,567 times.
To run unsloth/LFM2-8B-A1B-GGUF locally at a 4,096-token context, its quantized versions need between 4.42 GB (Q2_K, lowest quality) and 17.09 GB (BF16, highest quality) of memory, weights plus KV cache and a system margin included.
For most users the best balance is Q6_K_XL, needing about 7.96 GB. That means unsloth/LFM2-8B-A1B-GGUF fits entirely in the VRAM of a 6 GB GPU or larger, running fully on the GPU.
GGUF file size and memory by quantization
Compare real GGUF weight sizes, estimated KV cache and total memory for Q4, Q5, Q8 and every quantization published in this repository.
| Quant. | Bits | Quality | Weights | KV | Total | Speed~ | Verdict |
|---|---|---|---|---|---|---|---|
| Q2_K | 2.95 | Low | 2.87 GB | 0.75 GB | 4.42 GB | 139.5 t/s | Fits in VRAM |
| Q2_K_L | 2.95 | Low | 2.87 GB | 0.75 GB | 4.42 GB | 139.5 t/s | Fits in VRAM |
| Q2_K_XL | 2.99 | Low | 2.91 GB | 0.75 GB | 4.46 GB | 137.6 t/s | Fits in VRAM |
| Q3_K_S | 3.5 | Fair | 3.39 GB | 0.75 GB | 4.94 GB | 117.9 t/s | Fits in VRAM |
| Q3_K_XL | 3.53 | Fair | 3.42 GB | 0.75 GB | 4.97 GB | 116.8 t/s | Fits in VRAM |
| Q3_K_M | 3.83 | Fair | 3.72 GB | 0.75 GB | 5.27 GB | 107.4 t/s | Fits in VRAM |
| Q4_0 | 4.54 | Good | 4.41 GB | 0.75 GB | 5.96 GB | 90.7 t/s | Fits in VRAM |
| Q4_K_XL | 4.55 | Good | 4.42 GB | 0.75 GB | 5.97 GB | 90.5 t/s | Fits in VRAM |
| Q4_K_S | 4.56 | Good | 4.43 GB | 0.75 GB | 5.98 GB | 90.4 t/s | Fits in VRAM |
| Q4_K_M | 4.84 | Good | 4.7 GB | 0.75 GB | 6.25 GB | 85.1 t/s | Fits in VRAM |
| Q4_1 | 5.03 | Very good | 4.89 GB | 0.75 GB | 6.44 GB | 81.9 t/s | Fits in VRAM |
| Q5_K_S | 5.52 | Very good | 5.36 GB | 0.75 GB | 6.91 GB | 74.6 t/s | Fits in VRAM |
| Q5_K_M | 5.68 | Very good | 5.51 GB | 0.75 GB | 7.06 GB | 72.6 t/s | Fits in VRAM |
| Q5_K_XL | 5.68 | Very good | 5.51 GB | 0.75 GB | 7.06 GB | 72.6 t/s | Fits in VRAM |
| Q6_K | 6.57 | Excellent | 6.38 GB | 0.75 GB | 7.93 GB | 62.7 t/s | Fits in VRAM |
| Q6_K_XL | 6.6 | Excellent | 6.41 GB | 0.75 GB | 7.96 GB | 62.4 t/s | Fits in VRAM |
| Q8_0 | 8.51 | Excellent | 8.26 GB | 0.75 GB | 9.81 GB | 6.1 t/s | Offload |
| Q8_K_XL | 8.63 | Excellent | 8.38 GB | 0.75 GB | 9.93 GB | 6.0 t/s | Offload |
| BF16 | 16.01 | Excellent | 15.54 GB | 0.75 GB | 17.09 GB | 3.2 t/s | Offload |
KV cache computed from the model's exact architecture. Speed is a rough estimate bounded by memory bandwidth.
Frequently asked questions
How much VRAM do you need to run unsloth/LFM2-8B-A1B-GGUF?
You need about 5.98 GB of VRAM to run unsloth/LFM2-8B-A1B-GGUF entirely on the GPU using the Q4_K_S quantization (at a 4,096-token context). Smaller quantizations lower the requirement at the cost of quality.
Can I run unsloth/LFM2-8B-A1B-GGUF on an 8 GB GPU?
Yes. With 8 GB of VRAM you can run unsloth/LFM2-8B-A1B-GGUF fully on the GPU using Q6_K_XL (about 7.96 GB).
Can I run unsloth/LFM2-8B-A1B-GGUF on a 16 GB GPU?
Yes. With 16 GB of VRAM you can run unsloth/LFM2-8B-A1B-GGUF fully on the GPU using Q8_K_XL (about 9.93 GB).
Can I run unsloth/LFM2-8B-A1B-GGUF on a 24 GB GPU?
Yes. With 24 GB of VRAM you can run unsloth/LFM2-8B-A1B-GGUF fully on the GPU using BF16 (about 17.09 GB).
What is the best quantization for unsloth/LFM2-8B-A1B-GGUF?
If memory allows, higher bits-per-weight means better quality. A common sweet spot is a Q4_K_M or Q5_K_M quantization, which keeps most of the quality while roughly halving the memory versus 8-bit. Pick the highest quantization that still fits in your VRAM.