DeepSeek-R1-Distill-Qwen-1.5B GGUF size and VRAM requirements

License: apache-2.0 ⬇ 76,375 ❤ 149
Parameters1.78B
Context131,072

unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF is a compact reasoning-focused model with 1.78 billion parameters, built on the qwen2 architecture. It is released under the apache-2.0 license and has been downloaded 76,375 times.

To run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF locally at a 4,096-token context, its quantized versions need between 1.56 GB (IQ1_S, lowest quality) and 4.23 GB (BF16, highest quality) of memory, weights plus KV cache and a system margin included.

For most users the best balance is BF16, needing about 4.23 GB. That means unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF fits entirely in the VRAM of a 6 GB GPU or larger, running fully on the GPU.

→ Guide: How much VRAM do you need?

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 QualityWeights KVTotal Speed~Verdict
IQ1_S 3.15 Low 0.65 GB 0.11 GB 1.56 GB 613.4 t/s Fits in VRAM
IQ1_M 3.22 Low 0.67 GB 0.11 GB 1.58 GB 599.6 t/s Fits in VRAM
Q2_K 3.39 Fair 0.7 GB 0.11 GB 1.61 GB 570.5 t/s Fits in VRAM
IQ2_XXS 3.49 Fair 0.72 GB 0.11 GB 1.63 GB 554.2 t/s Fits in VRAM
IQ2_M 3.57 Fair 0.74 GB 0.11 GB 1.65 GB 542.2 t/s Fits in VRAM
Q2_K_L 3.64 Fair 0.75 GB 0.11 GB 1.66 GB 531.8 t/s Fits in VRAM
IQ3_XXS 3.97 Fair 0.82 GB 0.11 GB 1.73 GB 486.9 t/s Fits in VRAM
Q2_K_XL 3.99 Fair 0.82 GB 0.11 GB 1.73 GB 485.0 t/s Fits in VRAM
Q3_K_M 4.16 Fair 0.86 GB 0.11 GB 1.77 GB 464.6 t/s Fits in VRAM
Q3_K_XL 4.36 Good 0.9 GB 0.11 GB 1.81 GB 443.6 t/s Fits in VRAM
IQ4_XS 4.95 Good 1.02 GB 0.11 GB 1.93 GB 390.9 t/s Fits in VRAM
Q4_K_M 5.03 Very good 1.04 GB 0.11 GB 1.95 GB 384.4 t/s Fits in VRAM
Q4_K_XL 5.35 Very good 1.11 GB 0.11 GB 2.02 GB 361.1 t/s Fits in VRAM
Q5_K_M 5.79 Very good 1.2 GB 0.11 GB 2.11 GB 334.1 t/s Fits in VRAM
Q6_K 6.59 Excellent 1.36 GB 0.11 GB 2.27 GB 293.3 t/s Fits in VRAM
Q8_0 8.53 Excellent 1.76 GB 0.11 GB 2.67 GB 226.7 t/s Fits in VRAM
BF16 16.03 Excellent 3.32 GB 0.11 GB 4.23 GB 120.6 t/s Fits in VRAM

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/DeepSeek-R1-Distill-Qwen-1.5B-GGUF?

You need about 4.23 GB of VRAM to run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF entirely on the GPU using the BF16 quantization (at a 4,096-token context). Smaller quantizations lower the requirement at the cost of quality.

Can I run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF on an 8 GB GPU?

Yes. With 8 GB of VRAM you can run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF fully on the GPU using BF16 (about 4.23 GB).

Can I run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF on a 16 GB GPU?

Yes. With 16 GB of VRAM you can run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF fully on the GPU using BF16 (about 4.23 GB).

Can I run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF on a 24 GB GPU?

Yes. With 24 GB of VRAM you can run unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF fully on the GPU using BF16 (about 4.23 GB).

What is the best quantization for unsloth/DeepSeek-R1-Distill-Qwen-1.5B-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.