GPT-OSS 20B

21B MoE

OpenAI GPT-OSS

OpenAI open-weight MoE (21B total / 3.6B active), shipped natively in MXFP4 — ~12.8GB runs on a 16GB card with no quality tax. Only 3.6B active params means CPU-offload stays usable.

⬇ 8.1M HF downloads♥ 4913 likesopenai/gpt-oss-20b· stats from 8/14/2026

Consumer GPUMac / Apple SiliconCPU / VPS

131K

Max Context

Quant Variants

GGUF MXFP4

Best Quality

100.0%

Accuracy Retained

Calculate VRAM Hugging Face Compare

Quantization Variants

Per-quant VRAM, quality loss, and inference speed on RTX 4090

Measured = site benchmarks · Estimated = formula · Community = public reports

Format	Level	BPW	VRAM	PPL Loss	Speed	Source	Actions
GGUF	MXFP4	4.25	12.8 GB	0.0%	195 tok/s	Community	Calc HF
GGUF	Q8_0	5.1	13.8 GB	0.0%	178 tok/s	Estimated	Calc HF
GGUF	Q4_K_M	4.1	11.9 GB	1.4%	205 tok/s	Estimated	Calc HF

Similar models

Compare with GPT-OSS 120B

117B MoE

GPT-OSS 120B

OpenAI GPT-OSS

Pro GPUMac / Apple Silicon

58.5 GBmin VRAM·100.0%accuracy

The big GPT-OSS (117B total / 5.1B active). Native MXFP4 checkpoint is ~61GB — fits one 80GB card or a 128GB unified-memory Mac. Partial offload on 24GB consumer cards is slow but works.

30B-A3B

Qwen3 30B-A3B Instruct

Alibaba Qwen3

Consumer GPUMac / Apple Silicon

17.5 GBmin VRAM·98.9%accuracy

Qwen3 MoE with only 3B active params. Q4 ~19GB file; outperforms QwQ-32B on 16GB cards.

36B

Seed-OSS 36B Instruct

ByteDance Seed

Consumer GPUPro GPU

17.4 GBmin VRAM·98.7%accuracy

Dense 36B with a native 512K context. Q4 weights are ~22GB, which lands on a 24GB card or a 2×16GB split — but the context is the real cost: 512K of KV cache is roughly 128GB on its own, so budget context first and weights second.

32B

Qwen2.5 32B Instruct

Alibaba Qwen2.5

Consumer GPUPro GPU

16.4 GBmin VRAM·97.3%accuracy

Near-GPT-4 reasoning on a 24GB VRAM card (Q4_K_S). Groundbreaking value.

How to actually run this

Deployment guides for this model and this class of hardware.