moonshotai_Kimi-Linear-48B-A3B-Instruct GGUF size and VRAM requirements

License: mit ⬇ 4,744 ❤ 20
Parameters49.12B
Context1,048,576

bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF is a very large instruction-tuned chat model with 49.12 billion parameters, built on the kimi-linear architecture. It is released under the mit license and has been downloaded 4,744 times.

To run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF locally at a 4,096-token context, its quantized versions need between 4.32 GB (GGUF, lowest quality) and 95.72 GB (BF16, highest quality) of memory, weights plus KV cache and a system margin included.

For most users the best balance is GGUF, needing about 4.32 GB. That means bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF fits entirely in the VRAM of a 6 GB GPU or larger, running fully on the GPU.

Available GGUF quantizations for bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF include GGUF, IQ1_S, IQ1_M, IQ2_XXS, IQ2_XS, IQ2_S, IQ2_M, Q2_K, Q2_K_L, IQ3_XXS, IQ3_XS, Q3_K_S, IQ3_M, Q3_K_M, Q3_K_L, Q3_K_XL, IQ4_XS, IQ4_NL, Q4_0, Q4_K_S, Q4_K_M, Q4_K_L, Q4_1, Q5_K_S, Q5_K_M, Q5_K_L, Q6_K, Q6_K_L, Q8_0, BF16. The model supports a native context length of up to 1,048,576 tokens; a longer context grows the KV cache and the memory needed.

→ 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
GGUF 0.02 Very low 0.14 GB 3.38 GB 4.32 GB 2802.8 t/s Fits in VRAM
IQ1_S 1.71 Very low 9.77 GB 3.38 GB 13.95 GB 5.1 t/s Offload
IQ1_M 1.78 Very low 10.17 GB 3.38 GB 14.36 GB 4.9 t/s Offload
IQ2_XXS 1.98 Very low 11.3 GB 3.38 GB 15.48 GB 4.4 t/s Offload
IQ2_XS 2.26 Very low 12.94 GB 3.38 GB 17.12 GB 3.9 t/s Offload
IQ2_S 2.27 Very low 13.01 GB 3.38 GB 17.19 GB 3.8 t/s Offload
IQ2_M 2.57 Very low 14.71 GB 3.38 GB 18.89 GB 3.4 t/s Offload
Q2_K 2.86 Low 16.33 GB 3.38 GB 20.51 GB 3.1 t/s Offload
Q2_K_L 2.92 Low 16.67 GB 3.38 GB 20.85 GB 3.0 t/s Offload
IQ3_XXS 3.2 Low 18.3 GB 3.38 GB 22.48 GB 2.7 t/s Offload
IQ3_XS 3.33 Fair 19.04 GB 3.38 GB 23.22 GB 2.6 t/s Offload
Q3_K_S 3.52 Fair 20.12 GB 3.38 GB 24.3 GB Insufficient
IQ3_M 3.69 Fair 21.1 GB 3.38 GB 25.28 GB Insufficient
Q3_K_M 3.69 Fair 21.12 GB 3.38 GB 25.31 GB Insufficient
Q3_K_L 3.85 Fair 22.0 GB 3.38 GB 26.18 GB Insufficient
Q3_K_XL 3.9 Fair 22.3 GB 3.38 GB 26.49 GB Insufficient
IQ4_XS 4.31 Good 24.65 GB 3.38 GB 28.83 GB Insufficient
IQ4_NL 4.56 Good 26.05 GB 3.38 GB 30.23 GB Insufficient
Q4_0 4.63 Good 26.49 GB 3.38 GB 30.67 GB Insufficient
Q4_K_S 4.72 Good 26.99 GB 3.38 GB 31.17 GB Insufficient
Q4_K_M 4.9 Good 28.0 GB 3.38 GB 32.18 GB Insufficient
Q4_K_L 4.94 Good 28.26 GB 3.38 GB 32.44 GB Insufficient
Q4_1 5.04 Very good 28.85 GB 3.38 GB 33.03 GB Insufficient
Q5_K_S 5.54 Very good 31.68 GB 3.38 GB 35.86 GB Insufficient
Q5_K_M 5.72 Very good 32.69 GB 3.38 GB 36.87 GB Insufficient
Q5_K_L 5.75 Very good 32.9 GB 3.38 GB 37.09 GB Insufficient
Q6_K 6.59 Excellent 37.68 GB 3.38 GB 41.86 GB Insufficient
Q6_K_L 6.62 Excellent 37.85 GB 3.38 GB 42.03 GB Insufficient
Q8_0 8.51 Excellent 48.66 GB 3.38 GB 52.84 GB Insufficient
BF16 16.01 Excellent 91.54 GB 3.38 GB 95.72 GB Insufficient

KV cache estimated (architecture unavailable). Speed is a rough estimate bounded by memory bandwidth.

Frequently asked questions

What kind of model is bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF?

bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF is an instruction-tuned chat model with 49.12 billion parameters, based on the kimi-linear architecture. It is released under the mit license and distributed as GGUF files for local inference.

How much VRAM do you need to run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF?

You need about 4.32 GB of VRAM to run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF entirely on the GPU using the GGUF quantization (at a 4,096-token context). Smaller quantizations lower the requirement at the cost of quality.

Can I run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF on an 8 GB GPU?

Yes. With 8 GB of VRAM you can run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF fully on the GPU using GGUF (about 4.32 GB).

Can I run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF on a 16 GB GPU?

Yes. With 16 GB of VRAM you can run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF fully on the GPU using IQ2_XXS (about 15.48 GB).

Can I run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF on a 24 GB GPU?

Yes. With 24 GB of VRAM you can run bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF fully on the GPU using IQ3_XS (about 23.22 GB).

What context length does bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF support?

bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF supports a native context length of up to 1,048,576 tokens. A longer context grows the KV cache, so it increases the memory needed to run the model.

What is the best quantization for bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF?

For bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF, a strong default is Q4_K_M, which needs about 32.18 GB and keeps most of the quality while roughly halving the memory versus 8-bit. With VRAM to spare, Q5_K_M or Q6_K add a little more quality; if you are tight on memory, a smaller quantization still runs. Pick the highest quantization that fits your VRAM.