Qwen3-Coder-30B-A3B-Instruct-1M GGUF size and VRAM requirements
unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF is a very large code-focused language model with 30.53 billion parameters, built on the qwen3moe architecture. It is released under the apache-2.0 license and has been downloaded 12,008 times.
To run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF locally at a 4,096-token context, its quantized versions need between 3.58 GB (GGUF, lowest quality) and 60.37 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 3.58 GB. That means unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-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 |
|---|---|---|---|---|---|---|---|
| GGUF | 0.03 | Very low | 0.11 GB | 2.67 GB | 3.58 GB | 3519.6 t/s | Fits in VRAM |
| Q1_0 | 2.12 | Very low | 7.52 GB | 2.67 GB | 10.98 GB | 6.7 t/s | Offload |
| IQ1_S | 2.35 | Very low | 8.34 GB | 2.67 GB | 11.81 GB | 6.0 t/s | Offload |
| IQ1_M | 2.53 | Very low | 8.99 GB | 2.67 GB | 12.45 GB | 5.6 t/s | Offload |
| IQ2_XXS | 2.71 | Low | 9.63 GB | 2.67 GB | 13.1 GB | 5.2 t/s | Offload |
| IQ2_M | 2.84 | Low | 10.1 GB | 2.67 GB | 13.57 GB | 4.9 t/s | Offload |
| Q2_K | 2.95 | Low | 10.49 GB | 2.67 GB | 13.95 GB | 4.8 t/s | Offload |
| Q2_K_L | 2.97 | Low | 10.55 GB | 2.67 GB | 14.02 GB | 4.7 t/s | Offload |
| Q2_K_XL | 3.1 | Low | 11.0 GB | 2.67 GB | 14.47 GB | 4.5 t/s | Offload |
| IQ3_XXS | 3.37 | Fair | 11.99 GB | 2.67 GB | 15.45 GB | 4.2 t/s | Offload |
| Q3_K_S | 3.48 | Fair | 12.38 GB | 2.67 GB | 15.85 GB | 4.0 t/s | Offload |
| Q3_K_XL | 3.62 | Fair | 12.88 GB | 2.67 GB | 16.35 GB | 3.9 t/s | Offload |
| Q3_K_M | 3.85 | Fair | 13.7 GB | 2.67 GB | 17.17 GB | 3.6 t/s | Offload |
| IQ4_XS | 4.29 | Good | 15.25 GB | 2.67 GB | 18.72 GB | 3.3 t/s | Offload |
| IQ4_NL | 4.54 | Good | 16.12 GB | 2.67 GB | 19.59 GB | 3.1 t/s | Offload |
| Q4_0 | 4.55 | Good | 16.19 GB | 2.67 GB | 19.65 GB | 3.1 t/s | Offload |
| Q4_K_S | 4.57 | Good | 16.26 GB | 2.67 GB | 19.72 GB | 3.1 t/s | Offload |
| Q4_K_XL | 4.64 | Good | 16.48 GB | 2.67 GB | 19.94 GB | 3.0 t/s | Offload |
| Q4_K_M | 4.86 | Good | 17.28 GB | 2.67 GB | 20.75 GB | 2.9 t/s | Offload |
| Q4_1 | 5.03 | Very good | 17.87 GB | 2.67 GB | 21.34 GB | 2.8 t/s | Offload |
| Q5_K_S | 5.52 | Very good | 19.63 GB | 2.67 GB | 23.1 GB | 2.5 t/s | Offload |
| Q5_K_M | 5.69 | Very good | 20.23 GB | 2.67 GB | 23.7 GB | 2.5 t/s | Offload |
| Q5_K_XL | 5.7 | Very good | 20.25 GB | 2.67 GB | 23.71 GB | 2.5 t/s | Offload |
| Q6_K | 6.57 | Excellent | 23.37 GB | 2.67 GB | 26.84 GB | — | Insufficient |
| Q6_K_XL | 6.9 | Excellent | 24.53 GB | 2.67 GB | 28.0 GB | — | Insufficient |
| Q8_0 | 8.51 | Excellent | 30.25 GB | 2.67 GB | 33.72 GB | — | Insufficient |
| Q8_K_XL | 9.43 | Excellent | 33.52 GB | 2.67 GB | 36.98 GB | — | Insufficient |
| BF16 | 16.01 | Excellent | 56.9 GB | 2.67 GB | 60.37 GB | — | Insufficient |
KV cache estimated (architecture unavailable). Speed is a rough estimate bounded by memory bandwidth.
Frequently asked questions
How much VRAM do you need to run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF?
You need about 3.58 GB of VRAM to run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-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 unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF on an 8 GB GPU?
Yes. With 8 GB of VRAM you can run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF fully on the GPU using GGUF (about 3.58 GB).
Can I run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF on a 16 GB GPU?
Yes. With 16 GB of VRAM you can run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF fully on the GPU using Q3_K_S (about 15.85 GB).
Can I run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF on a 24 GB GPU?
Yes. With 24 GB of VRAM you can run unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-GGUF fully on the GPU using Q5_K_XL (about 23.71 GB).
What is the best quantization for unsloth/Qwen3-Coder-30B-A3B-Instruct-1M-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.