Radeon RX 7900 XTX vs RTX 3090 for local LLMs

Both run the same 48 of 69 models well. Speed on models that fit both is about the same.

AMD Radeon RX 7900 XTX

  • 24 GB VRAM
  • 960 GB/s bandwidth
  • 48/69 models run well
  • $999 at launch

NVIDIA RTX 3090

  • 24 GB VRAM
  • 936 GB/s bandwidth
  • 48/69 models run well
  • $1,499 at launch

Prices are US launch prices (MSRP) for comparison; street prices change daily. Links go to Amazon, where we earn a commission at no cost to you.

Every model, side by side

Best quant that fits and estimated speed at 8K context.

ModelRadeon RX 7900 XTXRTX 3090
Qwen3 0.6BFP16 · ~384 tok/sFP16 · ~374 tok/s
Gemma 3 1BFP16 · ~288 tok/sFP16 · ~281 tok/s
Llama 3.2 1BFP16 · ~232 tok/sFP16 · ~226 tok/s
Qwen3 1.7BFP16 · ~142 tok/sFP16 · ~138 tok/s
Llama 3.2 3BFP16 · ~90 tok/sFP16 · ~87 tok/s
Phi-3.5 Mini 3.8BFP16 · ~75 tok/sFP16 · ~74 tok/s
Phi-4 Mini 3.8BFP16 · ~75 tok/sFP16 · ~73 tok/s
Qwen3 4BFP16 · ~72 tok/sFP16 · ~70 tok/s
Gemma 3 4BFP16 · ~67 tok/sFP16 · ~65 tok/s
Mistral 7B v0.3FP16 · ~40 tok/sFP16 · ~39 tok/s
Qwen2.5 7BFP16 · ~38 tok/sFP16 · ~37 tok/s
Qwen2.5 Coder 7BFP16 · ~38 tok/sFP16 · ~37 tok/s
DeepSeek R1 Distill Qwen 7BFP16 · ~38 tok/sFP16 · ~37 tok/s
Llama 3.1 8BFP16 · ~36 tok/sFP16 · ~35 tok/s
Granite 3.3 8BFP16 · ~35 tok/sFP16 · ~34 tok/s
Qwen3 8BFP16 · ~35 tok/sFP16 · ~34 tok/s
DeepSeek R1 0528 Qwen3 8BFP16 · ~35 tok/sFP16 · ~34 tok/s
Gemma 2 9BFP16 · ~31 tok/sFP16 · ~30 tok/s
Gemma 4 12BQ8_0 · ~45 tok/sQ8_0 · ~44 tok/s
Mistral Nemo 12BQ8_0 · ~44 tok/sQ8_0 · ~43 tok/s
Gemma 3 12BQ8_0 · ~44 tok/sQ8_0 · ~43 tok/s
Qwen2.5 14BQ8_0 · ~37 tok/sQ8_0 · ~36 tok/s
Phi-4 14BQ8_0 · ~37 tok/sQ8_0 · ~36 tok/s
Phi-4 Reasoning 14BQ8_0 · ~37 tok/sQ8_0 · ~36 tok/s
Qwen3 14BQ8_0 · ~37 tok/sQ8_0 · ~36 tok/s
Qwen2.5 Coder 14BQ8_0 · ~37 tok/sQ8_0 · ~36 tok/s
DeepSeek R1 Distill Qwen 14BQ8_0 · ~37 tok/sQ8_0 · ~36 tok/s
gpt-oss 20BQ8_0 · ~151 tok/sQ8_0 · ~147 tok/s
Mistral Small 24BQ6_K · ~30 tok/sQ6_K · ~29 tok/s
Devstral Small 24BQ6_K · ~30 tok/sQ6_K · ~29 tok/s
Magistral Small 24BQ6_K · ~30 tok/sQ6_K · ~29 tok/s
Mistral Small 3.2 24BQ6_K · ~29 tok/sQ6_K · ~29 tok/s
Gemma 4 26B-A4BQ6_K · ~185 tok/sQ6_K · ~180 tok/s
Gemma 2 27BQ5_K_M · ~30 tok/sQ5_K_M · ~29 tok/s
Gemma 3 27BQ5_K_M · ~30 tok/sQ5_K_M · ~29 tok/s
Qwen3.8 27BQ6_K · ~25 tok/sQ6_K · ~25 tok/s
Qwen3.6 27BQ6_K · ~25 tok/sQ6_K · ~25 tok/s
Qwen3 30B-A3BQ5_K_M · ~245 tok/sQ5_K_M · ~239 tok/s
Qwen3 Coder 30B-A3BQ5_K_M · ~245 tok/sQ5_K_M · ~239 tok/s
GLM-4.7 Flash 30B-A3BQ5_K_M · ~270 tok/sQ5_K_M · ~263 tok/s
Gemma 4 31BQ4_K_M · ~30 tok/sQ4_K_M · ~30 tok/s
Nemotron 3 Nano 30B-A3BQ5_K_M · ~231 tok/sQ5_K_M · ~226 tok/s
Qwen2.5 32BQ4_K_M · ~29 tok/sQ4_K_M · ~28 tok/s
DeepSeek R1 Distill Qwen 32BQ4_K_M · ~29 tok/sQ4_K_M · ~28 tok/s
Qwen3 32BQ4_K_M · ~29 tok/sQ4_K_M · ~28 tok/s
QwQ 32BQ4_K_M · ~29 tok/sQ4_K_M · ~28 tok/s
Qwen2.5 Coder 32BQ4_K_M · ~29 tok/sQ4_K_M · ~28 tok/s
Qwen3.6 35B-A3BQ4_K_M · ~317 tok/sQ4_K_M · ~309 tok/s
Mixtral 8x7BQ2_K · ~119 tok/sQ2_K · ~116 tok/s
Llama 3.1 70BToo bigToo big
Llama 3.3 70BToo bigToo big
DeepSeek R1 Distill Llama 70BToo bigToo big
Qwen2.5 72BToo bigToo big
Qwen3 Coder Next 80B-A3BToo bigToo big
GLM-4.5-Air 106B-A12BToo bigToo big
Llama 4 Scout 17B-16EToo bigToo big
gpt-oss 120BToo bigToo big
Qwen3.8 Flash NextToo bigToo big
Qwen3 235B-A22BToo bigToo big
DeepSeek V4 Flash 284B-A13BToo bigToo big
GLM-5.3 Flash 320B-A18BToo bigToo big
MiniMax M3 428B-A23BToo bigToo big
GLM-5 744B-A40BToo bigToo big
GLM-5.3Too bigToo big
DeepSeek V4.1 FlashToo bigToo big
Mistral Large 4 1T-A52BToo bigToo big
DeepSeek V4 Pro 1.6T-A49BToo bigToo big
Qwen3.8 2.4T-A95BToo bigToo big
Kimi K3.1 Pro 2.8T-A104BToo bigToo big