vram.run Models Hardware Providers Cloud State of Inference
API provider data is live · Hardware & cloud pricing curated 2026-02-23

Radeon RX 6950 XT

AMD

16GB VRAM · 576 GB/s bandwidth · 47.3 FP16 TFLOPS · 335W TDP

The Radeon RX 6950 XT has 16GB of VRAM with 576 GB/s memory bandwidth and 47.3 TFLOPS FP16 compute. At Q4 quantization, it can comfortably run Gemma 3 4B (129 tok/s), Qwen 2.5 7B (68 tok/s), Llama 3.1 8B (64 tok/s). Models larger than ~27B parameters won't fit even at Q4. Electricity cost is approximately $36/month at 335W TDP.

What LLMs can you run?

ModelParamsQ4 WeightFitDecode
Gemma 3 4B4.0B2 GBcomfortable129 tok/s
Qwen 2.5 7B7.6B4 GBcomfortable68 tok/s
Llama 3.1 8B8.0B4 GBcomfortable64 tok/s
Mistral Small 24B24.0B12 GBtight21 tok/s
Gemma 3 27B27.4B14 GBwon't fit
Qwen 2.5 Coder 32B32.5B16 GBwon't fit
Llama 3.3 70B70.6B35 GBwon't fit
Qwen 2.5 72B72.7B36 GBwon't fit
Llama 3.1 405B405B202 GBwon't fit
DeepSeek R1 671B671B336 GBwon't fit

Similar GPUs

GPUVRAMBWTFLOPSTDP
Radeon PRO W770016GB576 GB/s63.9190W
Radeon RX 7900 GRE16GB576 GB/s92.0260W
Radeon RX 9070 GRE 16 GB16GB576 GB/s68.6220W
Radeon RX 770016GB622 GB/s53.2200W
Radeon RX 7800 XT16GB624 GB/s74.7263W

Compare with another GPU

Select another GPU to compare specs and model performance side by side.

Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0