AMD RX 6950 XT (16GB)
The AMD RX 6950 XT (16GB) provides 16 GB of memory with 576 GB/s bandwidth, supporting local inference via ROCm and Vulkan acceleration.
Memory
16 GB
Usable Memory Ceiling
16 GB
Bandwidth
576 GB/s
Max Inference Class
14B Models (Q4)
TDP
335W
VRAM Scaling by Context Length
Estimated memory requirements as context length increases. Curves crossing above the reference line indicate Out of Memory (OOM).
Precision:
Best Models to Run on AMD RX 6950 XT (16GB)
Inference throughput, memory compatibility, and generation speeds for open-weights LLMs on AMD RX 6950 XT (16GB).
Sort by:
Context:
KV Cache:
GPUs:
| Rank | Model | Parameters | Precision | TPS (Approx.) | TTFT | Can Run |
|---|---|---|---|---|---|---|
#48 | Qwen 3.8 27B | 27B | INT4 | 26 tok/s | ~8.4s | |
#93 | Qwen3.5-27B | 27B | INT4 | 26 tok/s | ~8.4s | |
#111 | MiMo V2 Flash | 15B (309B active) | INT4 | 3 tok/s | ~84.4s | |
#112 | Phi-4 Reasoning Plus | 14B | INT4 | 37 tok/s | ~3.7s | |
#117 | Gemma 4 26B A4B | 25.2B (3.8B active) | INT4 | 104 tok/s | ~1.4s | |
#125 | Qwen3.5-9B | 9B | INT4 Q8 | 70 tok/s 44 tok/s | ~2.4s ~2.4s | |
#128 | K2 Horizon 7B | 7B | INT4 Q8 FP16 | 84 tok/s 55 tok/s 28 tok/s | ~1.9s ~1.9s ~2.3s | |
#135 | Gemma 4 12B | 11.95B | INT4 Q8 | 42 tok/s 26 tok/s | ~3.2s ~3.8s | |
#140 | Qwen3.5-4B | 4B | INT4 Q8 FP16 | 119 tok/s 83 tok/s 49 tok/s | ~1.1s ~1.1s ~1.1s | |
#143 | GPT-OSS 20B | 21B (3.6B active) | INT4 | 51 tok/s | ~1.4s | |
#144 | GLM-4V | 9B | INT4 Q8 | 52 tok/s 35 tok/s | ~2.4s ~2.6s | |
#150 | NVIDIA Nemotron 3 Nano 30B-A3B | 3.5B (30B active) | INT4 Q8 FP16 | 25 tok/s 15 tok/s 8 tok/s | ~7.6s ~7.7s ~7.7s | |
#158 | Granite 4.2 8B | 8B | INT4 Q8 | 77 tok/s 49 tok/s | ~2.2s ~2.2s | |
#159 | Gemma 4 E4B | 8B | INT4 Q8 | 77 tok/s 49 tok/s | ~2.2s ~2.2s | |
#166 | Phi-4 | 14B | INT4 Q8 | 50 tok/s 27 tok/s | ~3.7s ~4.4s | |
#167 | Gemma 3 27B | 27B | INT4 | 26 tok/s | ~8.4s | |
#178 | OLMo 3 7B Think | 7B | INT4 Q8 | 62 tok/s 45 tok/s | ~1.9s ~1.9s | |
#179 | DeepSeek-R1 14B | 14B | INT4 Q8 | 51 tok/s 28 tok/s | ~3.7s ~4.4s | |
#183 | Ministral 3 14B | 14B | INT4 | 37 tok/s | ~3.7s | |
#186 | Gemma 3 12B | 12B | INT4 Q8 | 57 tok/s 33 tok/s | ~3.2s ~3.5s | |
#188 | Mistral-Small-2501 | 24B | INT4 | 30 tok/s | ~6.9s | |
#189 | Ministral 3 8B | 8B | INT4 Q8 | 57 tok/s 41 tok/s | ~2.2s ~2.2s | |
#191 | MiniCPM5-2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
#192 | Qwen3.5-2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
#193 | Gemma 2 27B | 27B | INT4 | 26 tok/s | ~8.4s | |
#194 | Qwen2.5-14B | 14B | INT4 Q8 | 50 tok/s 27 tok/s | ~3.7s ~4.4s | |
#200 | Gemma 4 E2B | 5.1B | INT4 Q8 FP16 | 103 tok/s 70 tok/s 40 tok/s | ~1.4s ~1.4s ~1.4s | |
#202 | DeepSeek-R1 8B | 8B | INT4 Q8 | 79 tok/s 50 tok/s | ~2.2s ~2.2s | |
#204 | Qwen3-4B | 4B | INT4 Q8 FP16 | 119 tok/s 83 tok/s 49 tok/s | ~1.1s ~1.1s ~1.1s | |
#205 | Ministral 3 3B | 3B | INT4 Q8 FP16 | 99 tok/s 79 tok/s 53 tok/s | ~856 ms ~859 ms ~865 ms | |
#207 | Gemma 2 9B | 9B | INT4 Q8 | 70 tok/s 44 tok/s | ~2.4s ~2.4s | |
#208 | Qwen3.5-0.8B | 0.8B | INT4 Q8 FP16 | 220 tok/s 189 tok/s 145 tok/s | ~246 ms ~246 ms ~248 ms | |
#209 | Qwen3-14B | 14B | INT4 Q8 | 50 tok/s 27 tok/s | ~3.7s ~4.4s | |
#210 | Gemma 3 4B | 4B | INT4 Q8 FP16 | 119 tok/s 83 tok/s 49 tok/s | ~1.1s ~1.1s ~1.1s | |
#211 | Qwen3-8B | 8B | INT4 Q8 | 77 tok/s 49 tok/s | ~2.2s ~2.2s | |
#213 | DeepSeek-R1 7B | 7B | INT4 Q8 FP16 | 86 tok/s 55 tok/s 29 tok/s | ~1.9s ~1.9s ~2.1s | |
#214 | Qwen3-1.7B | 1.7B | INT4 Q8 FP16 | 175 tok/s 137 tok/s 93 tok/s | ~509 ms ~511 ms ~514 ms | |
#215 | Llama 3.1 8B | 8B | INT4 Q8 | 77 tok/s 49 tok/s | ~2.2s ~2.2s | |
#216 | Gemma 3 270M | 0.27B | INT4 Q8 FP16 | 244 tok/s 231 tok/s 205 tok/s | ~86 ms ~86 ms ~87 ms | |
#218 | Qwen3-0.6B | 0.6B | INT4 Q8 FP16 | 234 tok/s 207 tok/s 165 tok/s | ~183 ms ~184 ms ~185 ms | |
#219 | Llama 3 8B | 8B | INT4 Q8 | 77 tok/s 49 tok/s | ~2.2s ~2.2s | |
#221 | Phi-4-Mini | 3.8B | INT4 Q8 FP16 | 122 tok/s 86 tok/s 51 tok/s | ~1.1s ~1.1s ~1.1s | |
#223 | Qwen2.5-7B | 7B | INT4 Q8 FP16 | 84 tok/s 55 tok/s 28 tok/s | ~1.9s ~1.9s ~2.3s | |
#224 | Ministral-8B-2410 | 8B | INT4 Q8 | 77 tok/s 49 tok/s | ~2.2s ~2.2s | |
#225 | DeepSeek-R1 1.5B | 1.5B | INT4 Q8 FP16 | 183 tok/s 146 tok/s 101 tok/s | ~452 ms ~453 ms ~456 ms | |
#226 | Llama 3.2 3B | 3B | INT4 Q8 FP16 | 138 tok/s 100 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
#228 | Phi-3-mini | 3.8B | INT4 Q8 FP16 | 122 tok/s 86 tok/s 51 tok/s | ~1.1s ~1.1s ~1.1s | |
#229 | Gemma 3n E2B IT | 6B (2B active) | INT4 Q8 FP16 | 99 tok/s 86 tok/s 60 tok/s | ~638 ms ~640 ms ~699 ms | |
#230 | Mistral-7B-Instruct-v0.1 | 7.3B | INT4 Q8 FP16 | 82 tok/s 53 tok/s 27 tok/s | ~2.0s ~2.0s ~2.3s | |
#231 | Llama 3.2 1B | 1B | INT4 Q8 FP16 | 207 tok/s 174 tok/s 128 tok/s | ~310 ms ~311 ms ~313 ms | |
#232 | Gemma 2 2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
#233 | Phi-3-medium | 14B | INT4 Q8 | 50 tok/s 27 tok/s | ~3.7s ~4.4s | |
#234 | OLMo 3 7B Instruct | 7B | INT4 Q8 | 62 tok/s 45 tok/s | ~1.9s ~1.9s | |
#236 | Qwen2-7B | 7B | INT4 Q8 FP16 | 84 tok/s 55 tok/s 28 tok/s | ~1.9s ~1.9s ~2.3s | |
#237 | Phi-3-small | 7B | INT4 Q8 FP16 | 84 tok/s 55 tok/s 28 tok/s | ~1.9s ~1.9s ~2.3s | |
#238 | Gemma 3 1B | 1B | INT4 Q8 FP16 | 207 tok/s 174 tok/s 128 tok/s | ~310 ms ~311 ms ~313 ms | |
#239 | Mistral-7B-Instruct-v0.2 | 7.3B | INT4 Q8 FP16 | 82 tok/s 53 tok/s 27 tok/s | ~2.0s ~2.0s ~2.3s | |
#241 | Gemma 1 7B | 7B | INT4 Q8 | 62 tok/s 45 tok/s | ~1.9s ~1.9s | |
#242 | Gemma 1 2B | 2B | INT4 Q8 FP16 | 172 tok/s 130 tok/s 85 tok/s | ~594 ms ~596 ms ~600 ms | |
#243 | ChatGLM3-6B | 6B | INT4 Q8 FP16 | 69 tok/s 51 tok/s 30 tok/s | ~1.6s ~1.6s ~1.8s | |
#244 | ChatGLM2-6B | 6B | INT4 Q8 FP16 | 69 tok/s 51 tok/s 30 tok/s | ~1.6s ~1.6s ~1.8s | |
#245 | ChatGLM-6B | 6B | INT4 Q8 FP16 | 69 tok/s 51 tok/s 30 tok/s | ~1.6s ~1.6s ~1.8s | |
Unranked Models (Sorted by Speed) | ||||||
- | DINOv3 ViT-L/16 | 0.3B | INT4 Q8 FP16 | 258 tok/s 241 tok/s 210 tok/s | ~95 ms ~95 ms ~96 ms | |
- | Qwen2-0.5B | 0.5B | INT4 Q8 FP16 | 242 tok/s 218 tok/s 178 tok/s | ~153 ms ~154 ms ~155 ms | |
- | Qwen2.5-0.5B | 0.5B | INT4 Q8 FP16 | 242 tok/s 218 tok/s 178 tok/s | ~153 ms ~154 ms ~155 ms | |
- | ERNIE-4.5-0.3B | 0.3B | INT4 Q8 FP16 | 239 tok/s 225 tok/s 198 tok/s | ~95 ms ~95 ms ~96 ms | |
- | ERNIE-4.5-0.3B-Base | 0.3B | INT4 Q8 FP16 | 239 tok/s 225 tok/s 198 tok/s | ~95 ms ~95 ms ~96 ms | |
- | Liquid D1-Omni 600M | 0.6B | INT4 Q8 FP16 | 234 tok/s 207 tok/s 165 tok/s | ~183 ms ~184 ms ~185 ms | |
- | PixArt-α XL 2 1024-MS | 0.61B | INT4 Q8 FP16 | 234 tok/s 206 tok/s 164 tok/s | ~186 ms ~187 ms ~188 ms | |
- | PixArt-Σ XL 2 1024-MS | 0.61B | INT4 Q8 FP16 | 234 tok/s 206 tok/s 164 tok/s | ~186 ms ~187 ms ~188 ms | |
- | Nemotron 3.5 Parse | 0.63B | INT4 Q8 FP16 | 232 tok/s 204 tok/s 162 tok/s | ~193 ms ~194 ms ~195 ms | |
- | SD 2.1 Turbo | 0.86B | INT4 Q8 FP16 | 216 tok/s 185 tok/s 139 tok/s | ~264 ms ~265 ms ~267 ms | |
- | Stable Diffusion 1.4 | 0.86B | INT4 Q8 FP16 | 216 tok/s 185 tok/s 139 tok/s | ~264 ms ~265 ms ~267 ms | |
- | Stable Diffusion 2.1 | 0.86B | INT4 Q8 FP16 | 216 tok/s 185 tok/s 139 tok/s | ~264 ms ~265 ms ~267 ms | |
- | Falcon-1B | 1B | INT4 Q8 FP16 | 214 tok/s 179 tok/s 131 tok/s | ~310 ms ~311 ms ~313 ms | |
- | Falcon3-1B | 1B | INT4 Q8 FP16 | 207 tok/s 174 tok/s 128 tok/s | ~310 ms ~311 ms ~313 ms | |
- | MiniCPM-V 4.7 1B | 1B | INT4 Q8 FP16 | 207 tok/s 174 tok/s 128 tok/s | ~310 ms ~311 ms ~313 ms | |
- | Kandinsky 2.2 Decoder | 1.22B | INT4 Q8 FP16 | 196 tok/s 161 tok/s 115 tok/s | ~373 ms ~374 ms ~376 ms | |
- | Wan 2.1 T2V 1.3B | 1.3B | INT4 Q8 FP16 | 192 tok/s 156 tok/s 110 tok/s | ~394 ms ~395 ms ~398 ms | |
- | Qwen2-1.5B | 1.5B | INT4 Q8 FP16 | 183 tok/s 146 tok/s 101 tok/s | ~452 ms ~453 ms ~456 ms | |
- | Qwen2.5-1.5B | 1.5B | INT4 Q8 FP16 | 183 tok/s 146 tok/s 101 tok/s | ~452 ms ~453 ms ~456 ms | |
- | SVD Image-to-Video XT | 1.5B | INT4 Q8 FP16 | 183 tok/s 146 tok/s 101 tok/s | ~452 ms ~453 ms ~456 ms | |
- | SVD Image-to-Video XT 1.1 | 1.5B | INT4 Q8 FP16 | 183 tok/s 146 tok/s 101 tok/s | ~452 ms ~453 ms ~456 ms | |
- | LFM3 1.7B | 1.7B | INT4 Q8 FP16 | 175 tok/s 137 tok/s 93 tok/s | ~509 ms ~511 ms ~514 ms | |
- | LFM3 1.7B Base | 1.7B | INT4 Q8 FP16 | 175 tok/s 137 tok/s 93 tok/s | ~509 ms ~511 ms ~514 ms | |
- | Spark X2.5 1.7B | 1.7B | INT4 Q8 FP16 | 175 tok/s 137 tok/s 93 tok/s | ~509 ms ~511 ms ~514 ms | |
- | Hy-MT2 1.8B | 1.8B | INT4 Q8 FP16 | 171 tok/s 134 tok/s 90 tok/s | ~536 ms ~538 ms ~542 ms | |
- | CogVideoX 2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
- | LFM3 Audio 2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
- | LFM3 Omni 2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
- | LFM3 VL 2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
- | Liquid D1 2B | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
- | LTX-Video 0.9.1 | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
- | Moondream 3.1 | 9B (2B active) | INT4 Q8 | 164 tok/s 126 tok/s | ~669 ms ~671 ms | |
- | Stable Diffusion 3 Medium | 2B | INT4 Q8 FP16 | 164 tok/s 126 tok/s 83 tok/s | ~595 ms ~596 ms ~601 ms | |
- | Kandinsky 3.0 | 2.1B | INT4 Q8 FP16 | 161 tok/s 123 tok/s 80 tok/s | ~621 ms ~622 ms ~627 ms | |
- | LTX-2 | 2.2B | INT4 Q8 FP16 | 158 tok/s 120 tok/s 78 tok/s | ~646 ms ~648 ms ~653 ms | |
- | SDXL 1.0 Refiner | 2.31B | INT4 Q8 FP16 | 155 tok/s 117 tok/s 75 tok/s | ~675 ms ~677 ms ~682 ms | |
- | Stable Diffusion 3.5 Medium | 2.5B | INT4 Q8 FP16 | 150 tok/s 112 tok/s 71 tok/s | ~724 ms ~726 ms ~732 ms | |
- | SDXL 1.0 Base | 2.56B | INT4 Q8 FP16 | 148 tok/s 110 tok/s 70 tok/s | ~740 ms ~742 ms ~747 ms | |
- | SDXL Turbo | 2.56B | INT4 Q8 FP16 | 148 tok/s 110 tok/s 70 tok/s | ~740 ms ~742 ms ~747 ms | |
- | Falcon-3B | 3B | INT4 Q8 FP16 | 144 tok/s 103 tok/s 63 tok/s | ~853 ms ~856 ms ~862 ms | |
- | CroissantLLM Base | 1.3B | INT4 Q8 FP16 | 142 tok/s 124 tok/s 93 tok/s | ~396 ms ~397 ms ~399 ms | |
- | Phi-1 | 1.3B | INT4 Q8 FP16 | 142 tok/s 124 tok/s 93 tok/s | ~396 ms ~397 ms ~399 ms | |
- | Phi-1.5 | 1.3B | INT4 Q8 FP16 | 142 tok/s 124 tok/s 93 tok/s | ~396 ms ~397 ms ~399 ms | |
- | DeepSeek-R1 3B | 3B | INT4 Q8 FP16 | 140 tok/s 101 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
- | Mellum 2.1 12B A2.5B Thinking | 12B (2.5B active) | INT4 Q8 | 140 tok/s 101 tok/s | ~834 ms ~908 ms | |
- | Falcon3-3B | 3B | INT4 Q8 FP16 | 138 tok/s 100 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
- | LFM2-VL 3B | 3B | INT4 Q8 FP16 | 138 tok/s 100 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
- | Liquid D1 3B | 3B | INT4 Q8 FP16 | 138 tok/s 100 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
- | Liquid D1-VL 3B | 3B | INT4 Q8 FP16 | 138 tok/s 100 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
- | Liquid LFM2-VL 3B D1 | 3B | INT4 Q8 FP16 | 138 tok/s 100 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
- | Qwen2.5-3B | 3B | INT4 Q8 FP16 | 138 tok/s 100 tok/s 62 tok/s | ~854 ms ~856 ms ~862 ms | |
- | LFM2.5 24B A2B | 24B (2.5B active) | INT4 | 124 tok/s | ~1.0s | |
- | FLUX.2 Klein 4B | 4B | INT4 Q8 FP16 | 119 tok/s 83 tok/s 49 tok/s | ~1.1s ~1.1s ~1.1s | |
- | FrogNano 4B | 4B | INT4 Q8 FP16 | 119 tok/s 83 tok/s 49 tok/s | ~1.1s ~1.1s ~1.1s | |
- | Mage-VL Cowork | 4B | INT4 Q8 FP16 | 119 tok/s 83 tok/s 49 tok/s | ~1.1s ~1.1s ~1.1s | |
- | Spark X2.5 4B | 4B | INT4 Q8 FP16 | 119 tok/s 83 tok/s 49 tok/s | ~1.1s ~1.1s ~1.1s | |
- | ERNIE-4.5-21B-A3B | 21B (3B active) | INT4 | 118 tok/s | ~1.1s | |
- | ERNIE-4.5-21B-A3B-Base | 21B (3B active) | INT4 | 118 tok/s | ~1.1s | |
- | EmbeddingGemma 2 | 2B | INT4 Q8 FP16 | 115 tok/s 97 tok/s 69 tok/s | ~597 ms ~599 ms ~603 ms | |
- | ERNIE-4.5-VL-28B-A3B | 28B (3B active) | INT4 | 108 tok/s | ~1.3s | |
- | ERNIE-4.5-VL-28B-A3B-Base | 28B (3B active) | INT4 | 108 tok/s | ~1.3s | |
- | CogVideoX 1.5 5B | 5B | INT4 Q8 FP16 | 105 tok/s 71 tok/s 41 tok/s | ~1.4s ~1.4s ~1.4s | |
- | CogVideoX 5B | 5B | INT4 Q8 FP16 | 105 tok/s 71 tok/s 41 tok/s | ~1.4s ~1.4s ~1.4s | |
- | Wan 2.2 TI2V 5B | 5B | INT4 Q8 FP16 | 105 tok/s 71 tok/s 41 tok/s | ~1.4s ~1.4s ~1.4s | |
- | Phi-2 | 2.7B | INT4 Q8 FP16 | 103 tok/s 84 tok/s 57 tok/s | ~779 ms ~781 ms ~787 ms | |
- | MaLLaM-3B | 3B | INT4 Q8 FP16 | 99 tok/s 79 tok/s 53 tok/s | ~856 ms ~859 ms ~865 ms | |
- | SmolLM3 3B | 3B | INT4 Q8 FP16 | 99 tok/s 79 tok/s 53 tok/s | ~856 ms ~859 ms ~865 ms | |
- | Falcon-7B | 7B | INT4 Q8 FP16 | 88 tok/s 56 tok/s 29 tok/s | ~1.9s ~1.9s ~2.1s | |
- | AuraFlow v0.2 | 6.8B | INT4 Q8 FP16 | 86 tok/s 56 tok/s 30 tok/s | ~1.8s ~1.9s ~2.0s | |
- | AuraFlow v0.3 | 6.8B | INT4 Q8 FP16 | 86 tok/s 56 tok/s 30 tok/s | ~1.8s ~1.9s ~2.0s | |
- | Falcon3-7B | 7B | INT4 Q8 FP16 | 84 tok/s 55 tok/s 28 tok/s | ~1.9s ~1.9s ~2.3s | |
- | Hy-MT2 7B | 7B | INT4 Q8 FP16 | 84 tok/s 55 tok/s 28 tok/s | ~1.9s ~1.9s ~2.3s | |
- | Mistral-7B-v0.1 | 7.3B | INT4 Q8 FP16 | 82 tok/s 53 tok/s 27 tok/s | ~2.0s ~2.0s ~2.3s | |
- | Youtu Parsing Omni | 8B | INT4 Q8 | 79 tok/s 50 tok/s | ~2.2s ~2.2s | |
- | Stable Diffusion 3.5 Large | 8.1B | INT4 Q8 | 76 tok/s 48 tok/s | ~2.2s ~2.2s | |
- | Stable Diffusion 3.5 Large Turbo | 8.1B | INT4 Q8 | 76 tok/s 48 tok/s | ~2.2s ~2.2s | |
- | FLUX.2 Klein 9B | 9B | INT4 Q8 | 70 tok/s 44 tok/s | ~2.4s ~2.4s | |
- | LensVLM 9B | 9B | INT4 Q8 | 70 tok/s 44 tok/s | ~2.4s ~2.4s | |
- | MiMo V2.6 Distill Qwen 9B | 9B | INT4 Q8 | 70 tok/s 44 tok/s | ~2.4s ~2.4s | |
- | ChatGLM3-6B-32K | 6B | INT4 Q8 FP16 | 69 tok/s 51 tok/s 30 tok/s | ~1.6s ~1.6s ~1.8s | |
- | Yi-6B | 6B | INT4 Q8 FP16 | 69 tok/s 51 tok/s 30 tok/s | ~1.6s ~1.6s ~1.8s | |
- | Falcon3-10B | 10B | INT4 Q8 | 65 tok/s 41 tok/s | ~2.7s ~2.7s | |
- | Mochi 1 Preview | 10B | INT4 Q8 | 65 tok/s 41 tok/s | ~2.7s ~2.7s | |
- | Falcon2-11B | 11B | INT4 Q8 | 63 tok/s 38 tok/s | ~2.9s ~3.0s | |
- | Hunyuan Lite | 7B | INT4 Q8 | 62 tok/s 45 tok/s | ~1.9s ~1.9s | |
- | Kimi-VL-A3B-Instruct | 16B (3B active) | INT4 | 62 tok/s | ~1.1s | |
- | OLMo 3 7B Base | 7B | INT4 Q8 | 62 tok/s 45 tok/s | ~1.9s ~1.9s | |
- | SEA-LION-7B | 7.1B | INT4 Q8 | 62 tok/s 45 tok/s | ~1.9s ~1.9s | |
- | SEA-LION-7B-Instruct | 7.1B | INT4 Q8 | 62 tok/s 45 tok/s | ~1.9s ~1.9s | |
- | FLUX.1 Canny Dev | 11.9B | INT4 Q8 | 57 tok/s 33 tok/s | ~3.2s ~3.5s | |
- | FLUX.1 Depth Dev | 11.9B | INT4 Q8 | 57 tok/s 33 tok/s | ~3.2s ~3.5s | |
- | FLUX.1 Dev | 11.9B | INT4 Q8 | 57 tok/s 33 tok/s | ~3.2s ~3.5s | |
- | FLUX.1 Fill Dev | 11.9B | INT4 Q8 | 57 tok/s 33 tok/s | ~3.2s ~3.5s | |
- | FLUX.1 Schnell | 11.9B | INT4 Q8 | 57 tok/s 33 tok/s | ~3.2s ~3.5s | |
- | Sahabat-AI-Llama3-8B-Instruct | 8B | INT4 Q8 | 57 tok/s 41 tok/s | ~2.2s ~2.2s | |
- | Typhoon-2-8B | 8B | INT4 Q8 | 57 tok/s 41 tok/s | ~2.2s ~2.2s | |
- | HunyuanVideo 1.5 720P T2V | 13B | INT4 Q8 | 53 tok/s 31 tok/s | ~3.5s ~3.8s | |
- | GLM-4-9B | 9B | INT4 Q8 | 52 tok/s 35 tok/s | ~2.4s ~2.6s | |
- | GLM-4-9B-Chat | 9B | INT4 Q8 | 52 tok/s 35 tok/s | ~2.4s ~2.6s | |
- | GLM-4-9B-Chat-1M | 9B | INT4 Q8 | 52 tok/s 35 tok/s | ~2.4s ~2.6s | |
- | Sahabat-AI-Gemma2-9B | 9.2B | INT4 Q8 | 52 tok/s 35 tok/s | ~2.5s ~2.7s | |
- | Sahabat-AI-Gemma2-9B-Instruct | 9.2B | INT4 Q8 | 52 tok/s 35 tok/s | ~2.5s ~2.7s | |
- | Yi-9B | 9B | INT4 Q8 | 52 tok/s 35 tok/s | ~2.4s ~2.6s | |
- | Wan 2.1 I2V 14B 720P | 14B | INT4 Q8 | 50 tok/s 27 tok/s | ~3.7s ~4.4s | |
- | Wan 2.1 T2V 14B | 14B | INT4 Q8 | 50 tok/s 27 tok/s | ~3.7s ~4.4s | |
- | Wan 2.2 T2V A14B | 14B | INT4 Q8 | 50 tok/s 27 tok/s | ~3.7s ~4.4s | |
- | FLUX.2 Dev | 15B | INT4 Q8 | 47 tok/s 26 tok/s | ~4.0s ~4.7s | |
- | GreenMind-14B-R1 | 14B | INT4 | 37 tok/s | ~3.7s | |
- | Arete 27B | 27B | INT4 | 26 tok/s | ~8.4s | |
- | AREX-2 | 27B | INT4 | 26 tok/s | ~8.4s | |
- | PPLX Decider v1 27B | 27B | INT4 | 26 tok/s | ~8.4s | |
- | PPLX Decider v1.1 27B | 27B | INT4 | 26 tok/s | ~8.4s | |
Inference Capacity
Maximum model parameter class runnable by weight precision.
Context:
4-bit
14B Models
8-bit
8B Models
16-bit
3B Models
Fine-Tuning Capacity
Hardware capacity for local training and adapter fine-tuning.
Context Length:
QLoRA
14B Models
LoRA
8B Models
Full Parameter Training
1B Models
Workload Recommendations
Guidance for optimal precision and operational ceilings on AMD RX 6950 XT (16GB).
Inference Sweet Spot
Optimal for 8B models across extended context (32k+). 14B models fit comfortably at Q4 precision with standard context.
Bandwidth & Speed Profile
With 576 GB/s aggregate bandwidth, batch size 1 inference operates in a memory-bandwidth bound regime. Generates approximately 128 tok/s on an 8B Q4 model and 15 tok/s on a 70B Q4 model.
Frequently Asked Questions
©2025 ApX Machine Learning