Skip to content
localmodel.run

Text model · Qwen3

Qwen3 4B requirements

Qwen3 family · 4B params · released 2025-04-29 · 23M Ollama pulls. Minimum to run at Q4_K_M: Apple M1 (8GB).

LicenseApache-2.0· Commercial OK↓ 10.5M/mo♥ 636on HuggingFace
Q4_K_M
2.5 GB
Q8_0
4.28 GB
Total @ Q4 (4k)
~3.8 GB
Context
32 k

Quantization sizes

GGUF quantson disk
QuantizationSize on disk
Q2_K1.7 GB est
Q3_K_M2 GB est
Q4_K_M (default)2.5 GB
Q5_K_M2.9 GB est
Q6_K3.3 GB est
Q8_04.28 GB
FP168 GB est

Lower quant = smaller and faster, slightly lower quality. Q4_K_M is the common default.

Run it

Ollama
$ ollama run qwen3:4b
llama.cpp
$ llama-cli -hf Qwen/Qwen3-4B-GGUF:Q4_K_M
LM Studio
$ lms get Qwen/Qwen3-4B-GGUF

Which devices can run Qwen3 4B?

FAQ

How much VRAM or RAM does Qwen3 4B need?

At Q4_K_M, Qwen3 4B needs about 3.8 GB (weights ~2.5 GB + KV cache + overhead) at a 4k context. At Q8_0 budget ~5.6 GB.

Can Qwen3 4B run on a laptop?

Yes, Qwen3 4B fits on a 16 GB machine at Q4_K_M and runs on Apple Silicon or a 12 GB+ GPU comfortably.

Can I use Qwen3 4B commercially?

Yes. Qwen3 4B is licensed Apache-2.0, which permits commercial use.

Claimed to rival Qwen2.5-72B-Instruct on several benchmarks. 32K native context (YaRN to 128K). Dual thinking/non-thinking mode. Excellent laptop-class edge model.

Sources

Memory figures are estimates. See methodology.