NVIDIA RTX 6000 Blackwell (96GB)
The NVIDIA RTX 6000 Blackwell (96GB) features 96 GB of dedicated VRAM and 1792 GB/s memory bandwidth, supporting CUDA and TensorRT-LLM runtimes for full-precision and quantized inference.
Memory
96 GB
Usable Memory Ceiling
96 GB
Bandwidth
1792 GB/s
Max Inference Class
70B+ Models (Q4)
TDP
500W
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 NVIDIA RTX 6000 Blackwell (96GB)
Inference throughput, memory compatibility, and generation speeds for open-weights LLMs on NVIDIA RTX 6000 Blackwell (96GB).
Sort by:
Context:
KV Cache:
GPUs:
| Rank | Model | Parameters | Precision | TPS (Approx.) | TTFT | Can Run |
|---|---|---|---|---|---|---|
#33 | Qwen 3.8 27B | 27B | INT4 Q8 FP16 | 78 tok/s 48 tok/s 26 tok/s | ~607 ms ~615 ms ~633 ms | |
#41 | Qwen 3.8 Flash | 176B (6B active) | INT4 | 194 tok/s | ~147 ms | |
#48 | Agnes 3.0 Flash | 33B | INT4 Q8 FP16 | 67 tok/s 40 tok/s 22 tok/s | ~737 ms ~746 ms ~768 ms | |
#61 | Sarvam-105B | 106B (10.3B active) | INT4 | 150 tok/s | ~311 ms | |
#72 | K2 Horizon MoVA 36B A4B | 36B (4B active) | INT4 Q8 FP16 | 214 tok/s 172 tok/s 118 tok/s | ~124 ms ~125 ms ~128 ms | |
#75 | Ling 3.0 Flash VL | 124B (5.5B active) | INT4 | 201 tok/s | ~223 ms | |
#83 | Qwen3.5-27B | 27B | INT4 Q8 FP16 | 78 tok/s 48 tok/s 26 tok/s | ~607 ms ~615 ms ~633 ms | |
#84 | Qwen3.5-122B-A10B | 122B (10B active) | INT4 | 125 tok/s | ~319 ms | |
#92 | Sarvam-30B | 32B (2.4B active) | INT4 Q8 FP16 | 245 tok/s 210 tok/s 158 tok/s | ~89 ms ~89 ms ~91 ms | |
#94 | Qwen3.6 35B A3B | 35B (3B active) | INT4 Q8 FP16 | 232 tok/s 194 tok/s 140 tok/s | ~102 ms ~103 ms ~105 ms | |
#96 | Muse Glimmer 30B | 30B | INT4 Q8 FP16 | 72 tok/s 44 tok/s 23 tok/s | ~672 ms ~681 ms ~701 ms | |
#100 | Phi-4 Reasoning Plus | 14B | INT4 Q8 FP16 | 98 tok/s 70 tok/s 43 tok/s | ~325 ms ~329 ms ~338 ms | |
#101 | MiMo V2 Flash | 15B (309B active) | INT4 Q8 FP16 | 9 tok/s 5 tok/s 2 tok/s | ~6.6s ~6.7s ~6.9s | |
#106 | GLM-4.5-Air | 106B (12B active) | INT4 | 53 tok/s | ~359 ms | |
#108 | Gemma 4 26B A4B | 25.2B (3.8B active) | INT4 Q8 FP16 | 222 tok/s 178 tok/s 123 tok/s | ~114 ms ~115 ms ~118 ms | |
#109 | Gemma 4 12B | 11.95B | INT4 Q8 FP16 | 108 tok/s 79 tok/s 49 tok/s | ~279 ms ~282 ms ~290 ms | |
#113 | Qwen3.5-9B | 9B | INT4 Q8 FP16 | 163 tok/s 113 tok/s 68 tok/s | ~211 ms ~214 ms ~220 ms | |
#114 | Qwen3.5-35B-A3B | 35B (3B active) | INT4 Q8 FP16 | 232 tok/s 194 tok/s 140 tok/s | ~102 ms ~103 ms ~105 ms | |
#114 | Ling 3.0 Flash | 124B (5.1B active) | INT4 | 206 tok/s | ~214 ms | |
#118 | K2 Horizon 7B | 7B | INT4 Q8 FP16 | 186 tok/s 134 tok/s 83 tok/s | ~167 ms ~169 ms ~174 ms | |
#126 | Qwen3 Next 80B A3B | 80B (79B active) | INT4 Q8 | 25 tok/s 14 tok/s | ~1.8s ~2.1s | |
#128 | NVIDIA Nemotron 3 Nano 30B-A3B | 3.5B (30B active) | INT4 Q8 FP16 | 69 tok/s 43 tok/s 23 tok/s | ~652 ms ~661 ms ~681 ms | |
#129 | Gemma 4 31B | 30.7B | INT4 Q8 FP16 | 71 tok/s 43 tok/s 23 tok/s | ~687 ms ~696 ms ~716 ms | |
#130 | GPT-OSS 120B | 117B (5.1B active) | INT4 | 53 tok/s | ~242 ms | |
#131 | Nemotron 3.5 Lightning | 30B (3B active) | INT4 Q8 FP16 | 128 tok/s 119 tok/s 97 tok/s | ~104 ms ~104 ms ~106 ms | |
#132 | Qwen3.5-4B | 4B | INT4 Q8 FP16 | 234 tok/s 183 tok/s 124 tok/s | ~101 ms ~102 ms ~105 ms | |
#133 | Qwen3-30B-A3B | 30B (3B active) | INT4 Q8 FP16 | 234 tok/s 195 tok/s 141 tok/s | ~100 ms ~101 ms ~103 ms | |
#136 | GPT-OSS 20B | 21B (3.6B active) | INT4 Q8 FP16 | 139 tok/s 125 tok/s 96 tok/s | ~110 ms ~111 ms ~113 ms | |
#144 | Mistral Medium 3.5 128B | 128B | INT4 | 15 tok/s | ~3.1s | |
#146 | Granite 4.2 8B | 8B | INT4 Q8 FP16 | 174 tok/s 123 tok/s 75 tok/s | ~189 ms ~191 ms ~197 ms | |
#150 | Gemma 4 E4B | 8B | INT4 Q8 FP16 | 174 tok/s 123 tok/s 75 tok/s | ~189 ms ~191 ms ~197 ms | |
#151 | Nemotron 3.5 Lightning 30B A3B | 30B (3B active) | INT4 Q8 FP16 | 234 tok/s 195 tok/s 141 tok/s | ~100 ms ~101 ms ~103 ms | |
#153 | Qwen3-32B | 32B | INT4 Q8 FP16 | 68 tok/s 41 tok/s 22 tok/s | ~715 ms ~725 ms ~746 ms | |
#154 | Qwen2.5-72B | 72B | INT4 Q8 | 34 tok/s 20 tok/s | ~1.6s ~1.6s | |
#155 | Mistral Large 3 | 41B (675B active) | INT4 Q8 FP16 | 4 tok/s 2 tok/s 1 tok/s | ~14.3s ~14.5s ~17.5s | |
#156 | GLM-4 | 32B | INT4 Q8 FP16 | 55 tok/s 36 tok/s 21 tok/s | ~719 ms ~728 ms ~749 ms | |
#157 | DeepSeek-R1 70B | 70B | INT4 Q8 | 36 tok/s 20 tok/s | ~1.5s ~1.6s | |
#158 | GLM-4V | 9B | INT4 Q8 FP16 | 129 tok/s 97 tok/s 62 tok/s | ~213 ms ~216 ms ~221 ms | |
#159 | Gemma 3 27B | 27B | INT4 Q8 FP16 | 78 tok/s 48 tok/s 26 tok/s | ~607 ms ~615 ms ~633 ms | |
#160 | Phi-4 | 14B | INT4 Q8 FP16 | 125 tok/s 82 tok/s 47 tok/s | ~322 ms ~327 ms ~336 ms | |
#161 | Llama 3.3 70B | 70B | INT4 Q8 | 35 tok/s 20 tok/s | ~1.5s ~1.6s | |
#162 | DeepSeek-R1 32B | 32B | INT4 Q8 FP16 | 70 tok/s 42 tok/s 22 tok/s | ~715 ms ~725 ms ~746 ms | |
#163 | Hunyuan Standard | 52B (389B active) | INT4 Q8 | 7 tok/s 4 tok/s | ~8.3s ~8.4s | |
#164 | Qwen2.5-32B | 32B | INT4 Q8 FP16 | 68 tok/s 41 tok/s 22 tok/s | ~715 ms ~725 ms ~746 ms | |
#168 | Magistral Small | 24B | INT4 Q8 FP16 | 68 tok/s 46 tok/s 27 tok/s | ~545 ms ~551 ms ~567 ms | |
#169 | OLMo 3 7B Think | 7B | INT4 Q8 FP16 | 149 tok/s 115 tok/s 75 tok/s | ~168 ms ~170 ms ~175 ms | |
#170 | OLMo 3.1 32B Instruct | 32B | INT4 Q8 FP16 | 55 tok/s 36 tok/s 21 tok/s | ~719 ms ~728 ms ~749 ms | |
#171 | Llama 3.1 70B | 70B | INT4 Q8 | 35 tok/s 20 tok/s | ~1.5s ~1.6s | |
#172 | DeepSeek-R1 14B | 14B | INT4 Q8 FP16 | 128 tok/s 83 tok/s 47 tok/s | ~322 ms ~326 ms ~336 ms | |
#173 | Ministral 3 14B | 14B | INT4 Q8 FP16 | 98 tok/s 70 tok/s 43 tok/s | ~325 ms ~329 ms ~338 ms | |
#174 | Mistral-Large-2407 | 123B | INT4 | 21 tok/s | ~2.7s | |
#175 | Llama 4 Scout | 109B (17B active) | INT4 | 97 tok/s | ~456 ms | |
#177 | Gemma 3 12B | 12B | INT4 Q8 FP16 | 138 tok/s 92 tok/s 53 tok/s | ~278 ms ~282 ms ~290 ms | |
#178 | Command A | 111B | INT4 | 18 tok/s | ~2.5s | |
#179 | Qwen2-72B | 72B | INT4 Q8 | 34 tok/s 20 tok/s | ~1.6s ~1.6s | |
#180 | OLMo 3 32B Base | 32B | INT4 Q8 FP16 | 55 tok/s 36 tok/s 21 tok/s | ~719 ms ~728 ms ~749 ms | |
#181 | Mistral-Small-2501 | 24B | INT4 Q8 FP16 | 85 tok/s 53 tok/s 29 tok/s | ~541 ms ~549 ms ~564 ms | |
#182 | Ministral 3 8B | 8B | INT4 Q8 FP16 | 139 tok/s 106 tok/s 68 tok/s | ~190 ms ~193 ms ~198 ms | |
#183 | Qwen3.5-2B | 2B | INT4 Q8 FP16 | 283 tok/s 243 tok/s 184 tok/s | ~57 ms ~58 ms ~59 ms | |
#184 | MiniCPM5-2B | 2B | INT4 Q8 FP16 | 283 tok/s 243 tok/s 184 tok/s | ~57 ms ~58 ms ~59 ms | |
#185 | Qwen2.5-14B | 14B | INT4 Q8 FP16 | 125 tok/s 82 tok/s 47 tok/s | ~322 ms ~327 ms ~336 ms | |
#186 | Gemma 2 27B | 27B | INT4 Q8 FP16 | 78 tok/s 48 tok/s 26 tok/s | ~607 ms ~615 ms ~633 ms | |
#187 | OLMo 3.1 32B Think | 32B | INT4 Q8 FP16 | 55 tok/s 36 tok/s 21 tok/s | ~719 ms ~728 ms ~749 ms | |
#188 | Gemma 4 E2B | 5.1B | INT4 Q8 FP16 | 214 tok/s 161 tok/s 105 tok/s | ~125 ms ~127 ms ~130 ms | |
#189 | Qwen3-4B | 4B | INT4 Q8 FP16 | 234 tok/s 183 tok/s 124 tok/s | ~101 ms ~102 ms ~105 ms | |
#190 | DeepSeek-R1 8B | 8B | INT4 Q8 FP16 | 177 tok/s 124 tok/s 75 tok/s | ~189 ms ~191 ms ~197 ms | |
#191 | Llama 3 70B | 70B | INT4 Q8 | 35 tok/s 20 tok/s | ~1.5s ~1.6s | |
#192 | Command R Plus | 104B | INT4 | 19 tok/s | ~2.3s | |
#193 | Ministral 3 3B | 3B | INT4 Q8 FP16 | 207 tok/s 177 tok/s 131 tok/s | ~80 ms ~81 ms ~83 ms | |
#195 | Qwen3-14B | 14B | INT4 Q8 FP16 | 125 tok/s 82 tok/s 47 tok/s | ~322 ms ~327 ms ~336 ms | |
#196 | Kimi Linear 48B A3B Instruct | 48B (3B active) | INT4 Q8 | 114 tok/s 107 tok/s | ~112 ms ~113 ms | |
#197 | Gemma 2 9B | 9B | INT4 Q8 FP16 | 163 tok/s 113 tok/s 68 tok/s | ~211 ms ~214 ms ~220 ms | |
#198 | Qwen2.5-7B | 7B | INT4 Q8 FP16 | 186 tok/s 134 tok/s 83 tok/s | ~167 ms ~169 ms ~174 ms | |
#199 | Qwen3-8B | 8B | INT4 Q8 FP16 | 174 tok/s 123 tok/s 75 tok/s | ~189 ms ~191 ms ~197 ms | |
#200 | Gemma 3 4B | 4B | INT4 Q8 FP16 | 234 tok/s 183 tok/s 124 tok/s | ~101 ms ~102 ms ~105 ms | |
#201 | DeepSeek-R1 7B | 7B | INT4 Q8 FP16 | 188 tok/s 135 tok/s 83 tok/s | ~167 ms ~169 ms ~174 ms | |
#202 | Llama 3.1 8B | 8B | INT4 Q8 FP16 | 174 tok/s 123 tok/s 75 tok/s | ~189 ms ~191 ms ~197 ms | |
#203 | Mixtral-8x22B-v0.1 | 176B (22B active) | INT4 | 73 tok/s | ~671 ms | |
#204 | Qwen3-1.7B | 1.7B | INT4 Q8 FP16 | 293 tok/s 256 tok/s 199 tok/s | ~50 ms ~50 ms ~51 ms | |
#205 | Command R | 35B | INT4 Q8 FP16 | 51 tok/s 34 tok/s 18 tok/s | ~784 ms ~794 ms ~884 ms | |
#206 | Gemma 3 270M | 0.27B | INT4 Q8 FP16 | 346 tok/s 337 tok/s 318 tok/s | ~14 ms ~14 ms ~14 ms | |
#208 | Llama 3 8B | 8B | INT4 Q8 FP16 | 174 tok/s 123 tok/s 75 tok/s | ~189 ms ~191 ms ~197 ms | |
#209 | Qwen3-0.6B | 0.6B | INT4 Q8 FP16 | 339 tok/s 320 tok/s 284 tok/s | ~22 ms ~22 ms ~23 ms | |
#210 | Phi-4-Mini | 3.8B | INT4 Q8 FP16 | 238 tok/s 188 tok/s 128 tok/s | ~96 ms ~98 ms ~100 ms | |
#211 | Mixtral-8x7B-v0.1 | 46.7B (7B active) | INT4 Q8 FP16 | 172 tok/s 127 tok/s 72 tok/s | ~192 ms ~194 ms ~232 ms | |
#212 | Ministral-8B-2410 | 8B | INT4 Q8 FP16 | 174 tok/s 123 tok/s 75 tok/s | ~189 ms ~191 ms ~197 ms | |
#213 | DeepSeek-R1 1.5B | 1.5B | INT4 Q8 FP16 | 300 tok/s 265 tok/s 210 tok/s | ~45 ms ~45 ms ~46 ms | |
#214 | Llama 3.2 3B | 3B | INT4 Q8 FP16 | 256 tok/s 209 tok/s 148 tok/s | ~79 ms ~80 ms ~82 ms | |
#215 | Qwen3.5-0.8B | 0.8B | INT4 Q8 FP16 | 329 tok/s 306 tok/s 263 tok/s | ~28 ms ~28 ms ~28 ms | |
#216 | Phi-3-mini | 3.8B | INT4 Q8 FP16 | 238 tok/s 188 tok/s 128 tok/s | ~96 ms ~98 ms ~100 ms | |
#217 | Gemma 3n E2B IT | 6B (2B active) | INT4 Q8 FP16 | 207 tok/s 188 tok/s 151 tok/s | ~61 ms ~62 ms ~63 ms | |
#218 | Mistral-7B-Instruct-v0.1 | 7.3B | INT4 Q8 FP16 | 182 tok/s 130 tok/s 80 tok/s | ~174 ms ~176 ms ~181 ms | |
#219 | Llama 3.2 1B | 1B | INT4 Q8 FP16 | 320 tok/s 292 tok/s 245 tok/s | ~33 ms ~33 ms ~34 ms | |
#220 | Gemma 2 2B | 2B | INT4 Q8 FP16 | 283 tok/s 243 tok/s 184 tok/s | ~57 ms ~58 ms ~59 ms | |
#221 | Phi-3-medium | 14B | INT4 Q8 FP16 | 125 tok/s 82 tok/s 47 tok/s | ~322 ms ~327 ms ~336 ms | |
#222 | OLMo 3 7B Instruct | 7B | INT4 Q8 FP16 | 149 tok/s 115 tok/s 75 tok/s | ~168 ms ~170 ms ~175 ms | |
#223 | Yi-34B | 34B | INT4 Q8 FP16 | 52 tok/s 35 tok/s 20 tok/s | ~762 ms ~772 ms ~794 ms | |
#224 | Qwen2-7B | 7B | INT4 Q8 FP16 | 186 tok/s 134 tok/s 83 tok/s | ~167 ms ~169 ms ~174 ms | |
#225 | Phi-3-small | 7B | INT4 Q8 FP16 | 186 tok/s 134 tok/s 83 tok/s | ~167 ms ~169 ms ~174 ms | |
#226 | Gemma 3 1B | 1B | INT4 Q8 FP16 | 320 tok/s 292 tok/s 245 tok/s | ~33 ms ~33 ms ~34 ms | |
#227 | Mistral-7B-Instruct-v0.2 | 7.3B | INT4 Q8 FP16 | 182 tok/s 130 tok/s 80 tok/s | ~174 ms ~176 ms ~181 ms | |
#228 | Falcon-180B | 180B | INT4 | 13 tok/s | ~4.6s | |
#229 | Gemma 1 7B | 7B | INT4 Q8 FP16 | 149 tok/s 115 tok/s 75 tok/s | ~168 ms ~170 ms ~175 ms | |
#230 | Gemma 1 2B | 2B | INT4 Q8 FP16 | 291 tok/s 248 tok/s 187 tok/s | ~57 ms ~57 ms ~59 ms | |
#231 | ChatGLM3-6B | 6B | INT4 Q8 FP16 | 160 tok/s 126 tok/s 84 tok/s | ~146 ms ~148 ms ~152 ms | |
#232 | ChatGLM2-6B | 6B | INT4 Q8 FP16 | 160 tok/s 126 tok/s 84 tok/s | ~146 ms ~148 ms ~152 ms | |
#233 | ChatGLM-6B | 6B | INT4 Q8 FP16 | 160 tok/s 126 tok/s 84 tok/s | ~146 ms ~148 ms ~152 ms | |
Unranked Models (Sorted by Speed) | ||||||
- | Qwen2-0.5B | 0.5B | INT4 Q8 FP16 | 344 tok/s 328 tok/s 296 tok/s | ~20 ms ~20 ms ~20 ms | |
- | Qwen2.5-0.5B | 0.5B | INT4 Q8 FP16 | 344 tok/s 328 tok/s 296 tok/s | ~20 ms ~20 ms ~20 ms | |
- | ERNIE-4.5-0.3B | 0.3B | INT4 Q8 FP16 | 342 tok/s 333 tok/s 313 tok/s | ~15 ms ~15 ms ~15 ms | |
- | ERNIE-4.5-0.3B-Base | 0.3B | INT4 Q8 FP16 | 342 tok/s 333 tok/s 313 tok/s | ~15 ms ~15 ms ~15 ms | |
- | Falcon-1B | 1B | INT4 Q8 FP16 | 325 tok/s 297 tok/s 248 tok/s | ~33 ms ~33 ms ~34 ms | |
- | Falcon3-1B | 1B | INT4 Q8 FP16 | 320 tok/s 292 tok/s 245 tok/s | ~33 ms ~33 ms ~34 ms | |
- | Qwen2-1.5B | 1.5B | INT4 Q8 FP16 | 300 tok/s 265 tok/s 210 tok/s | ~45 ms ~45 ms ~46 ms | |
- | Qwen2.5-1.5B | 1.5B | INT4 Q8 FP16 | 300 tok/s 265 tok/s 210 tok/s | ~45 ms ~45 ms ~46 ms | |
- | Spark X2.5 1.7B | 1.7B | INT4 Q8 FP16 | 293 tok/s 256 tok/s 199 tok/s | ~50 ms ~50 ms ~51 ms | |
- | Hy-MT2 1.8B | 1.8B | INT4 Q8 FP16 | 290 tok/s 251 tok/s 194 tok/s | ~52 ms ~53 ms ~54 ms | |
- | Falcon-3B | 3B | INT4 Q8 FP16 | 263 tok/s 213 tok/s 150 tok/s | ~79 ms ~80 ms ~82 ms | |
- | CroissantLLM Base | 1.3B | INT4 Q8 FP16 | 261 tok/s 240 tok/s 199 tok/s | ~41 ms ~41 ms ~42 ms | |
- | Phi-1 | 1.3B | INT4 Q8 FP16 | 261 tok/s 240 tok/s 199 tok/s | ~41 ms ~41 ms ~42 ms | |
- | Phi-1.5 | 1.3B | INT4 Q8 FP16 | 261 tok/s 240 tok/s 199 tok/s | ~41 ms ~41 ms ~42 ms | |
- | DeepSeek-R1 3B | 3B | INT4 Q8 FP16 | 258 tok/s 210 tok/s 148 tok/s | ~79 ms ~80 ms ~82 ms | |
- | Falcon3-3B | 3B | INT4 Q8 FP16 | 256 tok/s 209 tok/s 148 tok/s | ~79 ms ~80 ms ~82 ms | |
- | Qwen2.5-3B | 3B | INT4 Q8 FP16 | 256 tok/s 209 tok/s 148 tok/s | ~79 ms ~80 ms ~82 ms | |
- | ERNIE-4.5-21B-A3B | 21B (3B active) | INT4 Q8 FP16 | 240 tok/s 198 tok/s 143 tok/s | ~95 ms ~95 ms ~97 ms | |
- | ERNIE-4.5-21B-A3B-Base | 21B (3B active) | INT4 Q8 FP16 | 240 tok/s 198 tok/s 143 tok/s | ~95 ms ~95 ms ~97 ms | |
- | ERNIE-4.5-VL-28B-A3B | 28B (3B active) | INT4 Q8 FP16 | 235 tok/s 196 tok/s 141 tok/s | ~99 ms ~100 ms ~102 ms | |
- | ERNIE-4.5-VL-28B-A3B-Base | 28B (3B active) | INT4 Q8 FP16 | 235 tok/s 196 tok/s 141 tok/s | ~99 ms ~100 ms ~102 ms | |
- | Hy-MT2 30B A3B | 30B (3B active) | INT4 Q8 FP16 | 234 tok/s 195 tok/s 141 tok/s | ~100 ms ~101 ms ~103 ms | |
- | Spark X2.5 4B | 4B | INT4 Q8 FP16 | 234 tok/s 183 tok/s 124 tok/s | ~101 ms ~102 ms ~105 ms | |
- | Phi-2 | 2.7B | INT4 Q8 FP16 | 213 tok/s 185 tok/s 139 tok/s | ~73 ms ~74 ms ~76 ms | |
- | MaLLaM-3B | 3B | INT4 Q8 FP16 | 207 tok/s 177 tok/s 131 tok/s | ~80 ms ~81 ms ~83 ms | |
- | SmolLM3 3B | 3B | INT4 Q8 FP16 | 207 tok/s 177 tok/s 131 tok/s | ~80 ms ~81 ms ~83 ms | |
- | AliceAI Foundation 80B A3B Base | 80B (3B active) | INT4 Q8 | 206 tok/s 167 tok/s | ~135 ms ~146 ms | |
- | Falcon-7B | 7B | INT4 Q8 FP16 | 191 tok/s 136 tok/s 84 tok/s | ~167 ms ~169 ms ~174 ms | |
- | Falcon3-7B | 7B | INT4 Q8 FP16 | 186 tok/s 134 tok/s 83 tok/s | ~167 ms ~169 ms ~174 ms | |
- | Hy-MT2 7B | 7B | INT4 Q8 FP16 | 186 tok/s 134 tok/s 83 tok/s | ~167 ms ~169 ms ~174 ms | |
- | Mistral-7B-v0.1 | 7.3B | INT4 Q8 FP16 | 182 tok/s 130 tok/s 80 tok/s | ~174 ms ~176 ms ~181 ms | |
- | LensVLM 9B | 9B | INT4 Q8 FP16 | 163 tok/s 113 tok/s 68 tok/s | ~211 ms ~214 ms ~220 ms | |
- | MiMo V2.6 Distill Qwen 9B | 9B | INT4 Q8 FP16 | 163 tok/s 113 tok/s 68 tok/s | ~211 ms ~214 ms ~220 ms | |
- | ChatGLM3-6B-32K | 6B | INT4 Q8 FP16 | 160 tok/s 126 tok/s 84 tok/s | ~146 ms ~148 ms ~152 ms | |
- | Yi-6B | 6B | INT4 Q8 FP16 | 160 tok/s 126 tok/s 84 tok/s | ~146 ms ~148 ms ~152 ms | |
- | Falcon3-10B | 10B | INT4 Q8 FP16 | 154 tok/s 105 tok/s 62 tok/s | ~233 ms ~236 ms ~243 ms | |
- | Kimi-VL-A3B-Instruct | 16B (3B active) | INT4 Q8 FP16 | 154 tok/s 139 tok/s 109 tok/s | ~93 ms ~94 ms ~96 ms | |
- | Falcon2-11B | 11B | INT4 Q8 FP16 | 150 tok/s 100 tok/s 58 tok/s | ~255 ms ~259 ms ~266 ms | |
- | Hunyuan Lite | 7B | INT4 Q8 FP16 | 149 tok/s 115 tok/s 75 tok/s | ~168 ms ~170 ms ~175 ms | |
- | OLMo 3 7B Base | 7B | INT4 Q8 FP16 | 149 tok/s 115 tok/s 75 tok/s | ~168 ms ~170 ms ~175 ms | |
- | SEA-LION-7B | 7.1B | INT4 Q8 FP16 | 148 tok/s 114 tok/s 74 tok/s | ~171 ms ~173 ms ~177 ms | |
- | SEA-LION-7B-Instruct | 7.1B | INT4 Q8 FP16 | 148 tok/s 114 tok/s 74 tok/s | ~171 ms ~173 ms ~177 ms | |
- | Sahabat-AI-Llama3-8B-Instruct | 8B | INT4 Q8 FP16 | 139 tok/s 106 tok/s 68 tok/s | ~190 ms ~193 ms ~198 ms | |
- | Typhoon-2-8B | 8B | INT4 Q8 FP16 | 139 tok/s 106 tok/s 68 tok/s | ~190 ms ~193 ms ~198 ms | |
- | Laguna S 2.1 | 118B (8B active) | INT4 | 138 tok/s | ~275 ms | |
- | GLM-4-9B | 9B | INT4 Q8 FP16 | 129 tok/s 97 tok/s 62 tok/s | ~213 ms ~216 ms ~221 ms | |
- | GLM-4-9B-Chat | 9B | INT4 Q8 FP16 | 129 tok/s 97 tok/s 62 tok/s | ~213 ms ~216 ms ~221 ms | |
- | GLM-4-9B-Chat-1M | 9B | INT4 Q8 FP16 | 129 tok/s 97 tok/s 62 tok/s | ~213 ms ~216 ms ~221 ms | |
- | Yi-9B | 9B | INT4 Q8 FP16 | 129 tok/s 97 tok/s 62 tok/s | ~213 ms ~216 ms ~221 ms | |
- | Sahabat-AI-Gemma2-9B | 9.2B | INT4 Q8 FP16 | 128 tok/s 96 tok/s 61 tok/s | ~217 ms ~220 ms ~226 ms | |
- | Sahabat-AI-Gemma2-9B-Instruct | 9.2B | INT4 Q8 FP16 | 128 tok/s 96 tok/s 61 tok/s | ~217 ms ~220 ms ~226 ms | |
- | GreenMind-14B-R1 | 14B | INT4 Q8 FP16 | 98 tok/s 70 tok/s 43 tok/s | ~325 ms ~329 ms ~338 ms | |
- | Hunyuan A13B | 80B (13B active) | INT4 Q8 | 62 tok/s 47 tok/s | ~354 ms ~417 ms | |
- | Falcon-40B | 40B | INT4 Q8 FP16 | 58 tok/s 34 tok/s 17 tok/s | ~888 ms ~900 ms ~1.0s | |
- | Kimi-Dev-72B | 72B | INT4 Q8 | 28 tok/s 17 tok/s | ~1.6s ~1.8s | |
- | Typhoon-2-70B | 70B | INT4 Q8 | 28 tok/s 17 tok/s | ~1.6s ~1.7s | |
- | GLM-130B | 130B | INT4 | 15 tok/s | ~3.1s | |
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 NVIDIA RTX 6000 Blackwell (96GB).
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 1792 GB/s aggregate bandwidth, batch size 1 inference operates in a memory-bandwidth bound regime. Generates approximately 398 tok/s on an 8B Q4 model and 47 tok/s on a 70B Q4 model.
Frequently Asked Questions
©2025 ApX Machine Learning