NVIDIA H100 SXM5 80GB~259 tok/s (Q4)

openai/gpt-oss-safeguard-20b speed on NVIDIA H100 SXM5 80GB

Quantization-specific throughput and VRAM requirements for openai/gpt-oss-safeguard-20b running on NVIDIA H100 SXM5 80GB.

Speed Snapshot

Topline estimate from compatibility data

Modelopenai/gpt-oss-safeguard-20b

GPUNVIDIA H100 SXM5 80GB

Q4 speed259 tok/s

Q4 VRAM required11GB

Data Source

Calculation and benchmark status

Speed values come from the compatibility dataset (`estimatedTokensPerSec`) and are sorted by quantization.

For full verdict logic and alternate GPUs, see the canonical compatibility page.

Quantization Speed Table

Quantization	VRAM needed	VRAM available	Speed	Verdict
Q4	11GB	80GB	259 tok/s	✅ Fits
Q8	22GB	80GB	200 tok/s	✅ Fits
FP16	44GB	80GB	103 tok/s	✅ Fits