Qwen3 30B-A3B VRAM requirements

Running Qwen3 30B-A3B with experts in system RAM

llama.cpp's --n-cpu-moe N keeps the experts of the first N layers in RAM. For the measured Q4_K_M GGUF with 32K tokens of context, the smallest N that fits each card, and the speed with 89.6 GB/s of system RAM. Plan the offload for another GPU. The VRAM here can sit below the 22.7 GB of the worked example because it counts the file's measured bytes, keeps the input embedding (token_embd) in system RAM as llama.cpp does and adds a flat 1 GB of buffers, where the estimate above adds 0.5 GB plus 10% to the preset's weights.

In system RAM counts the offloaded experts plus the input embedding (token_embd), which stays in RAM even with --n-cpu-moe 0.

How much VRAM does Qwen3 30B-A3B need?

Qwen3 30B-A3B needs 20.2 GB at Q4_K_M, 34.6 GB at Q8_0 and 63.9 GB at FP16 with 8,192 tokens of context and one request. Each total below is the weights plus the FP16 KV cache and the runtime overhead (0.5 GB plus 10%), and each row opens the calculator with that setting. What the GGUF names mean.

Fine-tuning Qwen3 30B-A3B? Qwen3 30B-A3B VRAM for LoRA, QLoRA and full training.

Qwen3 30B-A3B GGUF files on Hugging Face

The 8 GGUF files llama.cpp picks for Qwen3 30B-A3B with -hf <repo>:<QUANT>, with the bytes the Hugging Face file list reported on 2026-09-29. Each total is that file plus the FP16 KV cache and 0.5 GB + 10%, one request. "vs estimate" compares the file with the bits-per-weight size the table above uses.

GB here is GiB (1024³ bytes), as everywhere on this site; Hugging Face shows the Q4_K_M file as 18.6 GB, in 10⁹ bytes.

Which Qwen3 30B-A3B GGUF fits a 16, 24, 32, 48 or 80 GB GPU or a Mac

From the file sizes above: the largest file that fits each machine with an 8K context and 0.5 GB left free, the longest context it then has room for, and the longest context for Q4_K_M (17.3 GB). One request, FP16 KV cache. Where nothing fits, the --n-cpu-moe table above keeps experts in system RAM.

Run Qwen3 30B-A3B with llama.cpp or Ollama

llama-server -hf ggml-org/Qwen3-30B-A3B-GGUF:Q4_K_M -c 38912

Qwen3-30B-A3B-Q4_K_M.gguf, 18.6 GB, from ggml-org/Qwen3-30B-A3B-GGUF (checked 2026-09-29). On a 24 GB RTX 3090 / 4090, Q4_K_M with -c 38912 takes 23.4 GB and leaves 584 MB.

ollama run qwen3:30b-a3b

Ollama library tag qwen3:30b-a3b; its quant is Ollama's choice. Same context: /set parameter num_ctx 38912.

Qwen3 30B-A3B at 8K, 32K and 40K (full) tokens of context

All 48 layers use full attention. Each extra token of context adds 96 KB of FP16 cache per request. The last column is the smallest setup here, one card or a group, that holds it at Q4_K_M. How the KV cache works.

How fast Qwen3 30B-A3B writes

Tokens per second for one request with 8,192 tokens of context, estimated from memory bandwidth and the parameters read per token. A dash means it does not fit on one card. Try other settings in the speed calculator.

Longest context on one GPU

How many tokens of context Qwen3 30B-A3B fits on a single card with one request, an FP16 KV cache and 0.5 GB left free. "Full" means the model's whole context window fits.

Model details

Parameters

30.5B (30,532,122,624)

Experts

128 routed experts, all loaded

Active per token

3.3B

Layers

All 48 layers use full attention

Attention cache

4 KV heads × 128

Context length

40,960 tokens

Published weights

56.9 GB (BF16)

On Hugging Face

Qwen/Qwen3-30B-A3B

Reported benchmarks

None published in the model card for GPQA, MMLU-Pro or LiveCodeBench.

Qwen3 30B-A3B next to similar-sized models

The three models here closest to it in parameter count, at Q4_K_M with 8,192 tokens of context.