Community performance summary

AMD Radeon RX 6900 XT

7 local models measured on this GPU configuration. Compare quantizations, configurations and the full observed performance range.

12 signed runs · 0 verified · 3 testers
Median prompt1226.03tok/s · 157.92 to 8088.69
Median generation89.66tok/s · 30.98 to 297.01
Configurations2Q4_K_M · Q8_0
Last observedAug 21RDNA 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.

Models on this GPU7 groups
Qwen3 VL 2B2 runsQwen3 30B-A3B1 runLlama 3.2 1B2 runsQwen3 4B2 runsGemma 4 12B3 runsGemma4 V2 Merged1 runGPT OSS 20B1 run
macOS

macOS 15.7.5 Sequoia · macOS 15.7.7 Sequoia · macOS 26.5.2 Tahoe

Recorded evidence

Inspect every run.

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

1 to 8 of 12 configurations

Verified benchmark Community submission
Model / quantHardwareArchitecturePromptGenerationSource
Llama 3.2 1BQ4_K_M · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
8088.69tok/s
297.01tok/s
Dolphins1972Jul 23, 2026
Llama 3.2 1BQ4_K_M · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
6224.67tok/s
206.18tok/s
ToshLLM communityAug 10, 2026
Qwen3 VL 2BQ8_0 · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
5211.36tok/s
157.49tok/s
Dolphins1972Jul 23, 2026
Qwen3 VL 2BQ8_0 · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
6342.45tok/s
129.18tok/s
ToshLLM communityAug 21, 2026
GPT OSS 20BQ4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
1652.01tok/s
118.87tok/s
Dolphins1972Jul 23, 2026
Qwen3 4BQ4_K_M · Dense · comparison
RX 6900 XT16 GB VRAM
RDNA 2
291.4tok/s
97.5tok/s
Community testerJul 16, 2026
Qwen3 4BQ4_K_M · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
2143.87tok/s
81.82tok/s
ToshLLM communityAug 10, 2026
Gemma 4 12BQ4_K_M · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
757.14tok/s
48.49tok/s
Dolphins1972Jul 23, 2026
Gemma 4 12BQ4_K_M · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
800.04tok/s
47.1tok/s
ToshLLM communityAug 1, 2026
Gemma4 V2 MergedQ4_K_M · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
799.02tok/s
45.89tok/s
ToshLLM communityAug 1, 2026
Qwen3 30B-A3BQ4_K_M · MoE · Signed ToshLLM app · ncmoe 16 · amd-gpu · Metal · comparison
RDNA 2
157.92tok/s
35.84tok/s
ToshLLM communityAug 21, 2026
Gemma 4 12BQ4_K_M · Dense · Signed ToshLLM app · amd-gpu · Metal · comparison
RDNA 2
763.1tok/s
30.98tok/s
ToshLLM communityAug 10, 2026