Community performance summary

GPT OSS 20B

4 GPU configurations measured with this model. Compare hardware differences, quantizations and every recorded result.

7 signed runs · 3 verified · 4 testers
Median prompt1038.29tok/s · 705.92 to 1652.01
Median generation94.59tok/s · 73.3 to 118.87
Configurations2MXFP4 · Q4_K_M
Last observedJul 27GCN / Vega + RDNA 2 · RDNA 1 · RDNA 2

What the results show

One clear view.
Every recorded run.

Measurements from comparable hardware and models are presented together. Medians show typical performance, while every signed run remains available below for inspection and sharing.

Recorded evidence

Inspect every run.

Open any result to review its exact hardware, model configuration, measurements and submission details.

1 to 8 of 7 configurations

Verified benchmark Community submission
Model / quantHardwareArchitecturePromptGenerationSource
GPT OSS 20BQ4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
1652.01tok/s
118.87tok/s
Dolphins1972Jul 23, 2026
GPT OSS 20BQ4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
GCN / Vega + RDNA 2
1334.01tok/s
110.45tok/s
gervaiscoquilJul 27, 2026
GPT OSS 20BQ4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
1623.43tok/s
106.16tok/s
Dolphins1972Jul 18, 2026
GPT OSS 20BQ4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
1038.29tok/s
94.59tok/s
ToshLLM LabJul 24, 2026
gpt-oss 20BQ4_K_M · MoE · Full VRAM · comparison
RX 6700 XT12 GB VRAM
RDNA 2
972tok/s
93tok/s
ToshLLM LabJul 15, 2026
GPT OSS 20BQ4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
1037.06tok/s
92.94tok/s
ToshLLM LabJul 22, 2026
GPT OSS 20BMXFP4 · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 1
705.92tok/s
73.3tok/s
ToshLLM communityJul 18, 2026