Instinct MI250X (128GB)
The Instinct MI250X (128GB) provides 128 GB of memory with 3276 GB/s bandwidth, supporting local inference via ROCm and Vulkan acceleration.
Memory
128 GB
Usable Memory Ceiling
128 GB
Bandwidth
3276 GB/s
Max Inference Class
70B+ Models (Q4)
TDP
560W
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 Instinct MI250X (128GB)
Inference throughput, memory compatibility, and generation speeds for open-weights LLMs on Instinct MI250X (128GB).
Sort by:
Context:
KV Cache:
GPUs:
| Rank | Model | Parameters | Precision | TPS (Approx.) | TTFT | Can Run |
|---|---|---|---|---|---|---|
#33 | Qwen 3.8 27B | 27B | INT4 Q8 FP16 | 136 tok/s 83 tok/s 45 tok/s | ~877 ms ~882 ms ~892 ms | |
#41 | Qwen 3.8 Flash | 176B (6B active) | INT4 | 360 tok/s | ~191 ms | |
#48 | Agnes 3.0 Flash | 33B | INT4 Q8 FP16 | 116 tok/s 70 tok/s 37 tok/s | ~1.1s ~1.1s ~1.1s | |
#61 | Sarvam-105B | 106B (10.3B active) | INT4 Q8 | 262 tok/s 170 tok/s | ~450 ms ~490 ms | |
#72 | K2 Horizon MoVA 36B A4B | 36B (4B active) | INT4 Q8 FP16 | 378 tok/s 301 tok/s 206 tok/s | ~174 ms ~175 ms ~176 ms | |
#75 | Ling 3.0 Flash VL | 124B (5.5B active) | INT4 Q8 | 354 tok/s 239 tok/s | ~321 ms ~378 ms | |
#83 | Qwen3.5-27B | 27B | INT4 Q8 FP16 | 136 tok/s 83 tok/s 45 tok/s | ~877 ms ~882 ms ~892 ms | |
#84 | Qwen3.5-122B-A10B | 122B (10B active) | INT4 Q8 | 219 tok/s 144 tok/s | ~461 ms ~543 ms | |
#92 | Sarvam-30B | 32B (2.4B active) | INT4 Q8 FP16 | 435 tok/s 371 tok/s 277 tok/s | ~122 ms ~123 ms ~123 ms | |
#94 | Qwen3.6 35B A3B | 35B (3B active) | INT4 Q8 FP16 | 411 tok/s 341 tok/s 245 tok/s | ~142 ms ~143 ms ~144 ms | |
#96 | Muse Glimmer 30B | 30B | INT4 Q8 FP16 | 125 tok/s 76 tok/s 41 tok/s | ~973 ms ~978 ms ~989 ms | |
#100 | Phi-4 Reasoning Plus | 14B | INT4 Q8 FP16 | 170 tok/s 122 tok/s 74 tok/s | ~464 ms ~466 ms ~472 ms | |
#101 | MiMo V2 Flash | 15B (309B active) | INT4 Q8 FP16 | 15 tok/s 8 tok/s 4 tok/s | ~9.6s ~9.6s ~9.7s | |
#106 | GLM-4.5-Air | 106B (12B active) | INT4 Q8 | 91 tok/s 72 tok/s | ~509 ms ~599 ms | |
#108 | Gemma 4 26B A4B | 25.2B (3.8B active) | INT4 Q8 FP16 | 392 tok/s 313 tok/s 216 tok/s | ~159 ms ~160 ms ~162 ms | |
#109 | Gemma 4 12B | 11.95B | INT4 Q8 FP16 | 189 tok/s 138 tok/s 85 tok/s | ~398 ms ~400 ms ~404 ms | |
#113 | Qwen3.5-9B | 9B | INT4 Q8 FP16 | 286 tok/s 198 tok/s 118 tok/s | ~301 ms ~302 ms ~306 ms | |
#114 | Qwen3.5-35B-A3B | 35B (3B active) | INT4 Q8 FP16 | 411 tok/s 341 tok/s 245 tok/s | ~142 ms ~143 ms ~144 ms | |
#114 | Ling 3.0 Flash | 124B (5.1B active) | INT4 Q8 | 364 tok/s 249 tok/s | ~309 ms ~364 ms | |
#118 | K2 Horizon 7B | 7B | INT4 Q8 FP16 | 326 tok/s 234 tok/s 144 tok/s | ~236 ms ~237 ms ~240 ms | |
#119 | Step 3.5 Flash | 196.81B (11B active) | INT4 | 194 tok/s | ~554 ms | |
#126 | Qwen3 Next 80B A3B | 80B (79B active) | INT4 Q8 | 44 tok/s 28 tok/s | ~2.5s ~2.6s | |
#128 | NVIDIA Nemotron 3 Nano 30B-A3B | 3.5B (30B active) | INT4 Q8 FP16 | 119 tok/s 74 tok/s 40 tok/s | ~942 ms ~947 ms ~959 ms | |
#129 | Gemma 4 31B | 30.7B | INT4 Q8 FP16 | 123 tok/s 74 tok/s 40 tok/s | ~995 ms ~1.0s ~1.0s | |
#130 | GPT-OSS 120B | 117B (5.1B active) | INT4 | 97 tok/s | ~312 ms | |
#131 | Nemotron 3.5 Lightning | 30B (3B active) | INT4 Q8 FP16 | 224 tok/s 209 tok/s 168 tok/s | ~142 ms ~142 ms ~143 ms | |
#132 | Qwen3.5-4B | 4B | INT4 Q8 FP16 | 414 tok/s 322 tok/s 216 tok/s | ~140 ms ~140 ms ~142 ms | |
#133 | Qwen3-30B-A3B | 30B (3B active) | INT4 Q8 FP16 | 414 tok/s 343 tok/s 246 tok/s | ~139 ms ~140 ms ~141 ms | |
#136 | GPT-OSS 20B | 21B (3.6B active) | INT4 Q8 FP16 | 243 tok/s 218 tok/s 168 tok/s | ~151 ms ~151 ms ~153 ms | |
#139 | Qwen3-235B-A22B | 235B (22B active) | INT4 | 122 tok/s | ~1.0s | |
#144 | Mistral Medium 3.5 128B | 128B | INT4 | 28 tok/s | ~4.1s | |
#146 | Granite 4.2 8B | 8B | INT4 Q8 FP16 | 305 tok/s 214 tok/s 130 tok/s | ~268 ms ~270 ms ~273 ms | |
#150 | Gemma 4 E4B | 8B | INT4 Q8 FP16 | 305 tok/s 214 tok/s 130 tok/s | ~268 ms ~270 ms ~273 ms | |
#151 | Nemotron 3.5 Lightning 30B A3B | 30B (3B active) | INT4 Q8 FP16 | 414 tok/s 343 tok/s 246 tok/s | ~139 ms ~140 ms ~141 ms | |
#153 | Qwen3-32B | 32B | INT4 Q8 FP16 | 119 tok/s 71 tok/s 38 tok/s | ~1.0s ~1.0s ~1.1s | |
#154 | Qwen2.5-72B | 72B | INT4 Q8 | 59 tok/s 34 tok/s | ~2.3s ~2.3s | |
#155 | Mistral Large 3 | 41B (675B active) | INT4 Q8 FP16 | 7 tok/s 4 tok/s 2 tok/s | ~20.9s ~21.0s ~21.3s | |
#156 | GLM-4 | 32B | INT4 Q8 FP16 | 95 tok/s 63 tok/s 36 tok/s | ~1.0s ~1.0s ~1.1s | |
#157 | DeepSeek-R1 70B | 70B | INT4 Q8 | 62 tok/s 35 tok/s | ~2.2s ~2.3s | |
#158 | GLM-4V | 9B | INT4 Q8 FP16 | 226 tok/s 169 tok/s 107 tok/s | ~302 ms ~303 ms ~307 ms | |
#159 | Gemma 3 27B | 27B | INT4 Q8 FP16 | 136 tok/s 83 tok/s 45 tok/s | ~877 ms ~882 ms ~892 ms | |
#160 | Phi-4 | 14B | INT4 Q8 FP16 | 218 tok/s 143 tok/s 81 tok/s | ~463 ms ~465 ms ~471 ms | |
#161 | Llama 3.3 70B | 70B | INT4 Q8 | 60 tok/s 35 tok/s | ~2.2s ~2.3s | |
#162 | DeepSeek-R1 32B | 32B | INT4 Q8 FP16 | 121 tok/s 72 tok/s 38 tok/s | ~1.0s ~1.0s ~1.1s | |
#163 | Hunyuan Standard | 52B (389B active) | INT4 Q8 FP16 | 12 tok/s 7 tok/s 3 tok/s | ~12.1s ~12.1s ~13.3s | |
#164 | Qwen2.5-32B | 32B | INT4 Q8 FP16 | 119 tok/s 71 tok/s 38 tok/s | ~1.0s ~1.0s ~1.1s | |
#168 | Magistral Small | 24B | INT4 Q8 FP16 | 117 tok/s 80 tok/s 46 tok/s | ~784 ms ~788 ms ~797 ms | |
#169 | OLMo 3 7B Think | 7B | INT4 Q8 FP16 | 260 tok/s 200 tok/s 131 tok/s | ~237 ms ~238 ms ~241 ms | |
#170 | OLMo 3.1 32B Instruct | 32B | INT4 Q8 FP16 | 95 tok/s 63 tok/s 36 tok/s | ~1.0s ~1.0s ~1.1s | |
#171 | Llama 3.1 70B | 70B | INT4 Q8 | 60 tok/s 35 tok/s | ~2.2s ~2.3s | |
#172 | DeepSeek-R1 14B | 14B | INT4 Q8 FP16 | 223 tok/s 145 tok/s 82 tok/s | ~463 ms ~465 ms ~471 ms | |
#173 | Ministral 3 14B | 14B | INT4 Q8 FP16 | 170 tok/s 122 tok/s 74 tok/s | ~464 ms ~466 ms ~472 ms | |
#174 | Mistral-Large-2407 | 123B | INT4 Q8 | 36 tok/s 18 tok/s | ~3.9s ~4.7s | |
#175 | Llama 4 Scout | 109B (17B active) | INT4 Q8 | 169 tok/s 108 tok/s | ~660 ms ~720 ms | |
#177 | Gemma 3 12B | 12B | INT4 Q8 FP16 | 241 tok/s 161 tok/s 93 tok/s | ~398 ms ~400 ms ~405 ms | |
#178 | Command A | 111B | INT4 Q8 | 31 tok/s 18 tok/s | ~3.6s ~4.2s | |
#179 | Qwen2-72B | 72B | INT4 Q8 | 59 tok/s 34 tok/s | ~2.3s ~2.3s | |
#180 | OLMo 3 32B Base | 32B | INT4 Q8 FP16 | 95 tok/s 63 tok/s 36 tok/s | ~1.0s ~1.0s ~1.1s | |
#181 | Mistral-Small-2501 | 24B | INT4 Q8 FP16 | 149 tok/s 92 tok/s 50 tok/s | ~782 ms ~786 ms ~796 ms | |
#182 | Ministral 3 8B | 8B | INT4 Q8 FP16 | 243 tok/s 184 tok/s 118 tok/s | ~269 ms ~270 ms ~273 ms | |
#183 | Qwen3.5-2B | 2B | INT4 Q8 FP16 | 504 tok/s 430 tok/s 324 tok/s | ~76 ms ~76 ms ~77 ms | |
#184 | MiniCPM5-2B | 2B | INT4 Q8 FP16 | 504 tok/s 430 tok/s 324 tok/s | ~76 ms ~76 ms ~77 ms | |
#185 | Qwen2.5-14B | 14B | INT4 Q8 FP16 | 218 tok/s 143 tok/s 81 tok/s | ~463 ms ~465 ms ~471 ms | |
#186 | Gemma 2 27B | 27B | INT4 Q8 FP16 | 136 tok/s 83 tok/s 45 tok/s | ~877 ms ~882 ms ~892 ms | |
#187 | OLMo 3.1 32B Think | 32B | INT4 Q8 FP16 | 95 tok/s 63 tok/s 36 tok/s | ~1.0s ~1.0s ~1.1s | |
#188 | Gemma 4 E2B | 5.1B | INT4 Q8 FP16 | 377 tok/s 283 tok/s 183 tok/s | ~175 ms ~176 ms ~178 ms | |
#189 | Qwen3-4B | 4B | INT4 Q8 FP16 | 414 tok/s 322 tok/s 216 tok/s | ~140 ms ~140 ms ~142 ms | |
#190 | DeepSeek-R1 8B | 8B | INT4 Q8 FP16 | 310 tok/s 217 tok/s 130 tok/s | ~268 ms ~269 ms ~273 ms | |
#191 | Llama 3 70B | 70B | INT4 Q8 | 60 tok/s 35 tok/s | ~2.2s ~2.3s | |
#192 | Command R Plus | 104B | INT4 Q8 | 33 tok/s 20 tok/s | ~3.3s ~3.7s | |
#193 | Ministral 3 3B | 3B | INT4 Q8 FP16 | 366 tok/s 312 tok/s 230 tok/s | ~108 ms ~108 ms ~110 ms | |
#195 | Qwen3-14B | 14B | INT4 Q8 FP16 | 218 tok/s 143 tok/s 81 tok/s | ~463 ms ~465 ms ~471 ms | |
#196 | Kimi Linear 48B A3B Instruct | 48B (3B active) | INT4 Q8 FP16 | 198 tok/s 187 tok/s 147 tok/s | ~153 ms ~153 ms ~167 ms | |
#197 | Gemma 2 9B | 9B | INT4 Q8 FP16 | 286 tok/s 198 tok/s 118 tok/s | ~301 ms ~302 ms ~306 ms | |
#198 | Qwen2.5-7B | 7B | INT4 Q8 FP16 | 326 tok/s 234 tok/s 144 tok/s | ~236 ms ~237 ms ~240 ms | |
#199 | Qwen3-8B | 8B | INT4 Q8 FP16 | 305 tok/s 214 tok/s 130 tok/s | ~268 ms ~270 ms ~273 ms | |
#200 | Gemma 3 4B | 4B | INT4 Q8 FP16 | 414 tok/s 322 tok/s 216 tok/s | ~140 ms ~140 ms ~142 ms | |
#201 | DeepSeek-R1 7B | 7B | INT4 Q8 FP16 | 332 tok/s 236 tok/s 145 tok/s | ~236 ms ~237 ms ~240 ms | |
#202 | Llama 3.1 8B | 8B | INT4 Q8 FP16 | 305 tok/s 214 tok/s 130 tok/s | ~268 ms ~270 ms ~273 ms | |
#203 | Mixtral-8x22B-v0.1 | 176B (22B active) | INT4 | 142 tok/s | ~832 ms | |
#204 | Qwen3-1.7B | 1.7B | INT4 Q8 FP16 | 523 tok/s 454 tok/s 350 tok/s | ~65 ms ~65 ms ~66 ms | |
#205 | Command R | 35B | INT4 Q8 FP16 | 89 tok/s 58 tok/s 33 tok/s | ~1.1s ~1.1s ~1.2s | |
#206 | Gemma 3 270M | 0.27B | INT4 Q8 FP16 | 620 tok/s 604 tok/s 569 tok/s | ~13 ms ~13 ms ~13 ms | |
#208 | Llama 3 8B | 8B | INT4 Q8 FP16 | 305 tok/s 214 tok/s 130 tok/s | ~268 ms ~270 ms ~273 ms | |
#209 | Qwen3-0.6B | 0.6B | INT4 Q8 FP16 | 608 tok/s 572 tok/s 506 tok/s | ~25 ms ~25 ms ~25 ms | |
#210 | Phi-4-Mini | 3.8B | INT4 Q8 FP16 | 422 tok/s 330 tok/s 223 tok/s | ~133 ms ~134 ms ~135 ms | |
#211 | Mixtral-8x7B-v0.1 | 46.7B (7B active) | INT4 Q8 FP16 | 302 tok/s 222 tok/s 139 tok/s | ~274 ms ~275 ms ~277 ms | |
#212 | Ministral-8B-2410 | 8B | INT4 Q8 FP16 | 305 tok/s 214 tok/s 130 tok/s | ~268 ms ~270 ms ~273 ms | |
#213 | DeepSeek-R1 1.5B | 1.5B | INT4 Q8 FP16 | 536 tok/s 471 tok/s 371 tok/s | ~58 ms ~58 ms ~59 ms | |
#214 | Llama 3.2 3B | 3B | INT4 Q8 FP16 | 455 tok/s 368 tok/s 259 tok/s | ~107 ms ~108 ms ~109 ms | |
#215 | Qwen3.5-0.8B | 0.8B | INT4 Q8 FP16 | 590 tok/s 546 tok/s 468 tok/s | ~33 ms ~33 ms ~33 ms | |
#216 | Phi-3-mini | 3.8B | INT4 Q8 FP16 | 422 tok/s 330 tok/s 223 tok/s | ~133 ms ~134 ms ~135 ms | |
#217 | Gemma 3n E2B IT | 6B (2B active) | INT4 Q8 FP16 | 366 tok/s 331 tok/s 264 tok/s | ~81 ms ~81 ms ~82 ms | |
#218 | Mistral-7B-Instruct-v0.1 | 7.3B | INT4 Q8 FP16 | 320 tok/s 228 tok/s 139 tok/s | ~245 ms ~247 ms ~250 ms | |
#219 | Llama 3.2 1B | 1B | INT4 Q8 FP16 | 572 tok/s 521 tok/s 434 tok/s | ~41 ms ~41 ms ~41 ms | |
#220 | Gemma 2 2B | 2B | INT4 Q8 FP16 | 504 tok/s 430 tok/s 324 tok/s | ~76 ms ~76 ms ~77 ms | |
#221 | Phi-3-medium | 14B | INT4 Q8 FP16 | 218 tok/s 143 tok/s 81 tok/s | ~463 ms ~465 ms ~471 ms | |
#222 | OLMo 3 7B Instruct | 7B | INT4 Q8 FP16 | 260 tok/s 200 tok/s 131 tok/s | ~237 ms ~238 ms ~241 ms | |
#223 | Yi-34B | 34B | INT4 Q8 FP16 | 91 tok/s 60 tok/s 34 tok/s | ~1.1s ~1.1s ~1.1s | |
#224 | Qwen2-7B | 7B | INT4 Q8 FP16 | 326 tok/s 234 tok/s 144 tok/s | ~236 ms ~237 ms ~240 ms | |
#225 | Phi-3-small | 7B | INT4 Q8 FP16 | 326 tok/s 234 tok/s 144 tok/s | ~236 ms ~237 ms ~240 ms | |
#226 | Gemma 3 1B | 1B | INT4 Q8 FP16 | 572 tok/s 521 tok/s 434 tok/s | ~41 ms ~41 ms ~41 ms | |
#227 | Mistral-7B-Instruct-v0.2 | 7.3B | INT4 Q8 FP16 | 320 tok/s 228 tok/s 139 tok/s | ~245 ms ~247 ms ~250 ms | |
#228 | Falcon-180B | 180B | INT4 | 26 tok/s | ~5.7s | |
#229 | Gemma 1 7B | 7B | INT4 Q8 FP16 | 260 tok/s 200 tok/s 131 tok/s | ~237 ms ~238 ms ~241 ms | |
#230 | Gemma 1 2B | 2B | INT4 Q8 FP16 | 518 tok/s 439 tok/s 329 tok/s | ~75 ms ~76 ms ~77 ms | |
#231 | ChatGLM3-6B | 6B | INT4 Q8 FP16 | 280 tok/s 220 tok/s 146 tok/s | ~205 ms ~206 ms ~208 ms | |
#232 | ChatGLM2-6B | 6B | INT4 Q8 FP16 | 280 tok/s 220 tok/s 146 tok/s | ~205 ms ~206 ms ~208 ms | |
#233 | ChatGLM-6B | 6B | INT4 Q8 FP16 | 280 tok/s 220 tok/s 146 tok/s | ~205 ms ~206 ms ~208 ms | |
Unranked Models (Sorted by Speed) | ||||||
- | Qwen2-0.5B | 0.5B | INT4 Q8 FP16 | 618 tok/s 587 tok/s 528 tok/s | ~21 ms ~21 ms ~21 ms | |
- | Qwen2.5-0.5B | 0.5B | INT4 Q8 FP16 | 618 tok/s 587 tok/s 528 tok/s | ~21 ms ~21 ms ~21 ms | |
- | ERNIE-4.5-0.3B | 0.3B | INT4 Q8 FP16 | 614 tok/s 597 tok/s 559 tok/s | ~14 ms ~14 ms ~14 ms | |
- | ERNIE-4.5-0.3B-Base | 0.3B | INT4 Q8 FP16 | 614 tok/s 597 tok/s 559 tok/s | ~14 ms ~14 ms ~14 ms | |
- | Falcon-1B | 1B | INT4 Q8 FP16 | 582 tok/s 529 tok/s 440 tok/s | ~41 ms ~41 ms ~41 ms | |
- | Falcon3-1B | 1B | INT4 Q8 FP16 | 572 tok/s 521 tok/s 434 tok/s | ~41 ms ~41 ms ~41 ms | |
- | Qwen2-1.5B | 1.5B | INT4 Q8 FP16 | 536 tok/s 471 tok/s 371 tok/s | ~58 ms ~58 ms ~59 ms | |
- | Qwen2.5-1.5B | 1.5B | INT4 Q8 FP16 | 536 tok/s 471 tok/s 371 tok/s | ~58 ms ~58 ms ~59 ms | |
- | Spark X2.5 1.7B | 1.7B | INT4 Q8 FP16 | 523 tok/s 454 tok/s 350 tok/s | ~65 ms ~65 ms ~66 ms | |
- | Hy-MT2 1.8B | 1.8B | INT4 Q8 FP16 | 517 tok/s 446 tok/s 341 tok/s | ~68 ms ~69 ms ~69 ms | |
- | Falcon-3B | 3B | INT4 Q8 FP16 | 467 tok/s 376 tok/s 263 tok/s | ~107 ms ~108 ms ~109 ms | |
- | CroissantLLM Base | 1.3B | INT4 Q8 FP16 | 463 tok/s 425 tok/s 351 tok/s | ~51 ms ~51 ms ~52 ms | |
- | Phi-1 | 1.3B | INT4 Q8 FP16 | 463 tok/s 425 tok/s 351 tok/s | ~51 ms ~51 ms ~52 ms | |
- | Phi-1.5 | 1.3B | INT4 Q8 FP16 | 463 tok/s 425 tok/s 351 tok/s | ~51 ms ~51 ms ~52 ms | |
- | DeepSeek-R1 3B | 3B | INT4 Q8 FP16 | 458 tok/s 370 tok/s 260 tok/s | ~107 ms ~108 ms ~109 ms | |
- | Falcon3-3B | 3B | INT4 Q8 FP16 | 455 tok/s 368 tok/s 259 tok/s | ~107 ms ~108 ms ~109 ms | |
- | Qwen2.5-3B | 3B | INT4 Q8 FP16 | 455 tok/s 368 tok/s 259 tok/s | ~107 ms ~108 ms ~109 ms | |
- | ERNIE-4.5-21B-A3B | 21B (3B active) | INT4 Q8 FP16 | 425 tok/s 350 tok/s 250 tok/s | ~131 ms ~131 ms ~132 ms | |
- | ERNIE-4.5-21B-A3B-Base | 21B (3B active) | INT4 Q8 FP16 | 425 tok/s 350 tok/s 250 tok/s | ~131 ms ~131 ms ~132 ms | |
- | ERNIE-4.5-VL-28B-A3B | 28B (3B active) | INT4 Q8 FP16 | 417 tok/s 345 tok/s 247 tok/s | ~137 ms ~138 ms ~139 ms | |
- | ERNIE-4.5-VL-28B-A3B-Base | 28B (3B active) | INT4 Q8 FP16 | 417 tok/s 345 tok/s 247 tok/s | ~137 ms ~138 ms ~139 ms | |
- | Hy-MT2 30B A3B | 30B (3B active) | INT4 Q8 FP16 | 414 tok/s 343 tok/s 246 tok/s | ~139 ms ~140 ms ~141 ms | |
- | Spark X2.5 4B | 4B | INT4 Q8 FP16 | 414 tok/s 322 tok/s 216 tok/s | ~140 ms ~140 ms ~142 ms | |
- | Phi-2 | 2.7B | INT4 Q8 FP16 | 377 tok/s 325 tok/s 243 tok/s | ~98 ms ~99 ms ~100 ms | |
- | MaLLaM-3B | 3B | INT4 Q8 FP16 | 366 tok/s 312 tok/s 230 tok/s | ~108 ms ~108 ms ~110 ms | |
- | SmolLM3 3B | 3B | INT4 Q8 FP16 | 366 tok/s 312 tok/s 230 tok/s | ~108 ms ~108 ms ~110 ms | |
- | AliceAI Foundation 80B A3B Base | 80B (3B active) | INT4 Q8 | 363 tok/s 310 tok/s | ~190 ms ~191 ms | |
- | Falcon-7B | 7B | INT4 Q8 FP16 | 336 tok/s 238 tok/s 146 tok/s | ~236 ms ~237 ms ~240 ms | |
- | Falcon3-7B | 7B | INT4 Q8 FP16 | 326 tok/s 234 tok/s 144 tok/s | ~236 ms ~237 ms ~240 ms | |
- | Hy-MT2 7B | 7B | INT4 Q8 FP16 | 326 tok/s 234 tok/s 144 tok/s | ~236 ms ~237 ms ~240 ms | |
- | Mistral-7B-v0.1 | 7.3B | INT4 Q8 FP16 | 320 tok/s 228 tok/s 139 tok/s | ~245 ms ~247 ms ~250 ms | |
- | LensVLM 9B | 9B | INT4 Q8 FP16 | 286 tok/s 198 tok/s 118 tok/s | ~301 ms ~302 ms ~306 ms | |
- | MiMo V2.6 Distill Qwen 9B | 9B | INT4 Q8 FP16 | 286 tok/s 198 tok/s 118 tok/s | ~301 ms ~302 ms ~306 ms | |
- | ChatGLM3-6B-32K | 6B | INT4 Q8 FP16 | 280 tok/s 220 tok/s 146 tok/s | ~205 ms ~206 ms ~208 ms | |
- | Yi-6B | 6B | INT4 Q8 FP16 | 280 tok/s 220 tok/s 146 tok/s | ~205 ms ~206 ms ~208 ms | |
- | Kimi-VL-A3B-Instruct | 16B (3B active) | INT4 Q8 FP16 | 270 tok/s 244 tok/s 191 tok/s | ~127 ms ~127 ms ~128 ms | |
- | Falcon3-10B | 10B | INT4 Q8 FP16 | 269 tok/s 184 tok/s 108 tok/s | ~333 ms ~335 ms ~339 ms | |
- | Falcon2-11B | 11B | INT4 Q8 FP16 | 262 tok/s 175 tok/s 101 tok/s | ~365 ms ~367 ms ~371 ms | |
- | Hunyuan Lite | 7B | INT4 Q8 FP16 | 260 tok/s 200 tok/s 131 tok/s | ~237 ms ~238 ms ~241 ms | |
- | OLMo 3 7B Base | 7B | INT4 Q8 FP16 | 260 tok/s 200 tok/s 131 tok/s | ~237 ms ~238 ms ~241 ms | |
- | SEA-LION-7B | 7.1B | INT4 Q8 FP16 | 259 tok/s 199 tok/s 129 tok/s | ~240 ms ~241 ms ~244 ms | |
- | SEA-LION-7B-Instruct | 7.1B | INT4 Q8 FP16 | 259 tok/s 199 tok/s 129 tok/s | ~240 ms ~241 ms ~244 ms | |
- | Sahabat-AI-Llama3-8B-Instruct | 8B | INT4 Q8 FP16 | 243 tok/s 184 tok/s 118 tok/s | ~269 ms ~270 ms ~273 ms | |
- | Typhoon-2-8B | 8B | INT4 Q8 FP16 | 243 tok/s 184 tok/s 118 tok/s | ~269 ms ~270 ms ~273 ms | |
- | Laguna S 2.1 | 118B (8B active) | INT4 Q8 | 242 tok/s 174 tok/s | ~395 ms ~431 ms | |
- | GLM-4-9B | 9B | INT4 Q8 FP16 | 226 tok/s 169 tok/s 107 tok/s | ~302 ms ~303 ms ~307 ms | |
- | GLM-4-9B-Chat | 9B | INT4 Q8 FP16 | 226 tok/s 169 tok/s 107 tok/s | ~302 ms ~303 ms ~307 ms | |
- | GLM-4-9B-Chat-1M | 9B | INT4 Q8 FP16 | 226 tok/s 169 tok/s 107 tok/s | ~302 ms ~303 ms ~307 ms | |
- | Yi-9B | 9B | INT4 Q8 FP16 | 226 tok/s 169 tok/s 107 tok/s | ~302 ms ~303 ms ~307 ms | |
- | Sahabat-AI-Gemma2-9B | 9.2B | INT4 Q8 FP16 | 224 tok/s 167 tok/s 105 tok/s | ~308 ms ~310 ms ~313 ms | |
- | Sahabat-AI-Gemma2-9B-Instruct | 9.2B | INT4 Q8 FP16 | 224 tok/s 167 tok/s 105 tok/s | ~308 ms ~310 ms ~313 ms | |
- | GreenMind-14B-R1 | 14B | INT4 Q8 FP16 | 170 tok/s 122 tok/s 74 tok/s | ~464 ms ~466 ms ~472 ms | |
- | Hunyuan A13B | 80B (13B active) | INT4 Q8 | 108 tok/s 90 tok/s | ~503 ms ~505 ms | |
- | Falcon-40B | 40B | INT4 Q8 FP16 | 101 tok/s 59 tok/s 31 tok/s | ~1.3s ~1.3s ~1.3s | |
- | Typhoon-2-70B | 70B | INT4 Q8 | 49 tok/s 31 tok/s | ~2.3s ~2.3s | |
- | Kimi-Dev-72B | 72B | INT4 Q8 | 48 tok/s 30 tok/s | ~2.3s ~2.3s | |
- | GLM-130B | 130B | INT4 | 27 tok/s | ~4.2s | |
Inference Capacity
Maximum model parameter class runnable by weight precision.
Context:
4-bit
70B+ Models
8-bit
70B Models
16-bit
32B Models
Fine-Tuning Capacity
Hardware capacity for local training and adapter fine-tuning.
Context Length:
QLoRA
70B+ Models
LoRA
70B Models
Full Parameter Training
7B-8B Models
Workload Recommendations
Guidance for optimal precision and operational ceilings on Instinct MI250X (128GB).
Inference Sweet Spot
Suitable for 70B parameter models at Q4/Q8 with extended 32k context windows, or high-throughput batching of smaller dense architectures.
Bandwidth & Speed Profile
With 3276 GB/s aggregate bandwidth, batch size 1 inference operates in a memory-bandwidth bound regime. Generates approximately 728 tok/s on an 8B Q4 model and 86 tok/s on a 70B Q4 model.
Frequently Asked Questions
©2025 ApX Machine Learning