NVIDIA RTX 5000 Ada Generation (32GB)
The NVIDIA RTX 5000 Ada Generation (32GB) features 32 GB of dedicated VRAM and 576 GB/s memory bandwidth, supporting CUDA and TensorRT-LLM runtimes for full-precision and quantized inference.
Memory
32 GB
Usable Memory Ceiling
32 GB
Bandwidth
576 GB/s
Max Inference Class
32B Models (Q4)
TDP
250W
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 5000 Ada Generation (32GB)
Inference throughput, memory compatibility, and generation speeds for open-weights LLMs on NVIDIA RTX 5000 Ada Generation (32GB).
Sort by:
Context:
KV Cache:
GPUs:
| Rank | Model | Parameters | Precision | TPS (Approx.) | TTFT | Can Run |
|---|---|---|---|---|---|---|
#36 | Qwen 3.8 27B | 27B | INT4 Q8 | 29 tok/s 16 tok/s | ~1.2s ~1.3s | |
#80 | K2 Horizon MoVA 36B A4B | 36B (4B active) | INT4 | 113 tok/s | ~233 ms | |
#86 | Qwen3.5-27B | 27B | INT4 Q8 | 29 tok/s 16 tok/s | ~1.2s ~1.3s | |
#93 | Sarvam-30B | 32B (2.4B active) | INT4 | 143 tok/s | ~164 ms | |
#99 | Muse Glimmer 30B | 30B | INT4 Q8 | 27 tok/s 14 tok/s | ~1.3s ~1.5s | |
#102 | Phi-4 Reasoning Plus | 14B | INT4 Q8 FP16 | 38 tok/s 26 tok/s 13 tok/s | ~622 ms ~634 ms ~773 ms | |
#103 | Qwen3.6 35B A3B | 35B (3B active) | INT4 | 130 tok/s | ~191 ms | |
#105 | MiMo V2 Flash | 15B (309B active) | INT4 Q8 | 3 tok/s 2 tok/s | ~12.7s ~13.0s | |
#110 | Gemma 4 26B A4B | 25.2B (3.8B active) | INT4 Q8 | 120 tok/s 81 tok/s | ~214 ms ~235 ms | |
#120 | Qwen3.5-35B-A3B | 35B (3B active) | INT4 | 130 tok/s | ~191 ms | |
#121 | Qwen3.5-9B | 9B | INT4 Q8 FP16 | 75 tok/s 46 tok/s 25 tok/s | ~401 ms ~409 ms ~428 ms | |
#127 | K2 Horizon 7B | 7B | INT4 Q8 FP16 | 90 tok/s 57 tok/s 31 tok/s | ~315 ms ~322 ms ~336 ms | |
#130 | Gemma 4 12B | 11.95B | INT4 Q8 FP16 | 44 tok/s 30 tok/s 16 tok/s | ~534 ms ~544 ms ~615 ms | |
#136 | Qwen3.5-4B | 4B | INT4 Q8 FP16 | 132 tok/s 89 tok/s 51 tok/s | ~188 ms ~191 ms ~199 ms | |
#139 | GPT-OSS 20B | 21B (3.6B active) | INT4 Q8 | 60 tok/s 49 tok/s | ~209 ms ~229 ms | |
#140 | GLM-4V | 9B | INT4 Q8 FP16 | 55 tok/s 38 tok/s 22 tok/s | ~406 ms ~414 ms ~432 ms | |
#145 | NVIDIA Nemotron 3 Nano 30B-A3B | 3.5B (30B active) | INT4 Q8 FP16 | 25 tok/s 15 tok/s 8 tok/s | ~1.3s ~1.3s ~1.3s | |
#150 | Qwen3-30B-A3B | 30B (3B active) | INT4 Q8 | 132 tok/s 88 tok/s | ~187 ms ~221 ms | |
#152 | Granite 4.2 8B | 8B | INT4 Q8 FP16 | 82 tok/s 51 tok/s 28 tok/s | ~358 ms ~365 ms ~381 ms | |
#157 | Gemma 4 E4B | 8B | INT4 Q8 FP16 | 82 tok/s 51 tok/s 28 tok/s | ~358 ms ~365 ms ~381 ms | |
#158 | Mistral Large 3 | 41B (675B active) | INT4 | 1 tok/s | ~32.4s | |
#161 | GLM-4 | 32B | INT4 | 20 tok/s | ~1.4s | |
#163 | Gemma 3 27B | 27B | INT4 Q8 | 29 tok/s 16 tok/s | ~1.2s ~1.3s | |
#164 | Phi-4 | 14B | INT4 Q8 FP16 | 52 tok/s 31 tok/s 16 tok/s | ~615 ms ~628 ms ~711 ms | |
#166 | Qwen3-32B | 32B | INT4 | 25 tok/s | ~1.4s | |
#168 | DeepSeek-R1 32B | 32B | INT4 Q8 | 26 tok/s 13 tok/s | ~1.4s ~1.6s | |
#169 | Gemma 4 31B | 30.7B | INT4 Q8 | 26 tok/s 13 tok/s | ~1.3s ~1.6s | |
#171 | Magistral Small | 24B | INT4 Q8 | 25 tok/s 15 tok/s | ~1.0s ~1.2s | |
#172 | Qwen2.5-32B | 32B | INT4 | 25 tok/s | ~1.4s | |
#174 | OLMo 3.1 32B Instruct | 32B | INT4 | 20 tok/s | ~1.4s | |
#176 | DeepSeek-R1 14B | 14B | INT4 Q8 FP16 | 54 tok/s 32 tok/s 16 tok/s | ~614 ms ~627 ms ~710 ms | |
#177 | Ministral 3 14B | 14B | INT4 Q8 FP16 | 38 tok/s 26 tok/s 13 tok/s | ~622 ms ~634 ms ~773 ms | |
#180 | Gemma 3 12B | 12B | INT4 Q8 FP16 | 59 tok/s 36 tok/s 19 tok/s | ~530 ms ~541 ms ~565 ms | |
#184 | Mistral-Small-2501 | 24B | INT4 Q8 | 33 tok/s 18 tok/s | ~1.0s ~1.1s | |
#185 | Ministral 3 8B | 8B | INT4 Q8 FP16 | 60 tok/s 42 tok/s 25 tok/s | ~362 ms ~369 ms ~385 ms | |
#186 | Gemma 2 27B | 27B | INT4 Q8 | 29 tok/s 16 tok/s | ~1.2s ~1.3s | |
#187 | OLMo 3 7B Think | 7B | INT4 Q8 FP16 | 66 tok/s 47 tok/s 28 tok/s | ~319 ms ~325 ms ~340 ms | |
#188 | Qwen3.5-2B | 2B | INT4 Q8 FP16 | 189 tok/s 140 tok/s 89 tok/s | ~103 ms ~105 ms ~109 ms | |
#189 | MiniCPM5-2B | 2B | INT4 Q8 FP16 | 189 tok/s 140 tok/s 89 tok/s | ~103 ms ~105 ms ~109 ms | |
#190 | Qwen2.5-14B | 14B | INT4 Q8 FP16 | 52 tok/s 31 tok/s 16 tok/s | ~615 ms ~628 ms ~711 ms | |
#191 | OLMo 3.1 32B Think | 32B | INT4 | 20 tok/s | ~1.4s | |
#193 | Gemma 4 E2B | 5.1B | INT4 Q8 FP16 | 113 tok/s 74 tok/s 42 tok/s | ~234 ms ~239 ms ~249 ms | |
#195 | Ministral 3 3B | 3B | INT4 Q8 FP16 | 107 tok/s 84 tok/s 56 tok/s | ~148 ms ~150 ms ~157 ms | |
#196 | DeepSeek-R1 8B | 8B | INT4 Q8 FP16 | 84 tok/s 52 tok/s 28 tok/s | ~357 ms ~365 ms ~381 ms | |
#198 | Gemma 2 9B | 9B | INT4 Q8 FP16 | 75 tok/s 46 tok/s 25 tok/s | ~401 ms ~409 ms ~428 ms | |
#199 | Nemotron 3.5 Lightning 30B A3B | 30B (3B active) | INT4 Q8 | 132 tok/s 88 tok/s | ~187 ms ~221 ms | |
#200 | Kimi Linear 48B A3B Instruct | 48B (3B active) | INT4 | 41 tok/s | ~250 ms | |
#201 | Qwen3-4B | 4B | INT4 Q8 FP16 | 132 tok/s 89 tok/s 51 tok/s | ~188 ms ~191 ms ~199 ms | |
#202 | Qwen3-14B | 14B | INT4 Q8 FP16 | 52 tok/s 31 tok/s 16 tok/s | ~615 ms ~628 ms ~711 ms | |
#203 | Gemma 3 4B | 4B | INT4 Q8 FP16 | 132 tok/s 89 tok/s 51 tok/s | ~188 ms ~191 ms ~199 ms | |
#204 | DeepSeek-R1 7B | 7B | INT4 Q8 FP16 | 93 tok/s 58 tok/s 32 tok/s | ~315 ms ~321 ms ~336 ms | |
#205 | OLMo 3 32B Base | 32B | INT4 | 20 tok/s | ~1.4s | |
#206 | Qwen3.5-0.8B | 0.8B | INT4 Q8 FP16 | 267 tok/s 223 tok/s 164 tok/s | ~46 ms ~47 ms ~48 ms | |
#207 | Llama 3.1 8B | 8B | INT4 Q8 FP16 | 82 tok/s 51 tok/s 28 tok/s | ~358 ms ~365 ms ~381 ms | |
#207 | Qwen3-8B | 8B | INT4 Q8 FP16 | 82 tok/s 51 tok/s 28 tok/s | ~358 ms ~365 ms ~381 ms | |
#210 | Nemotron 3.5 Lightning | 30B (3B active) | INT4 | 54 tok/s | ~198 ms | |
#211 | Command R | 35B | INT4 | 18 tok/s | ~1.5s | |
#212 | Llama 3 8B | 8B | INT4 Q8 FP16 | 82 tok/s 51 tok/s 28 tok/s | ~358 ms ~365 ms ~381 ms | |
#213 | Qwen3-1.7B | 1.7B | INT4 Q8 FP16 | 203 tok/s 154 tok/s 101 tok/s | ~89 ms ~91 ms ~94 ms | |
#214 | Gemma 3 270M | 0.27B | INT4 Q8 FP16 | 303 tok/s 283 tok/s 245 tok/s | ~20 ms ~21 ms ~21 ms | |
#216 | Qwen3-0.6B | 0.6B | INT4 Q8 FP16 | 288 tok/s 249 tok/s 191 tok/s | ~36 ms ~37 ms ~38 ms | |
#217 | Phi-4-Mini | 3.8B | INT4 Q8 FP16 | 136 tok/s 92 tok/s 54 tok/s | ~179 ms ~183 ms ~190 ms | |
#218 | Qwen2.5-7B | 7B | INT4 Q8 FP16 | 90 tok/s 57 tok/s 31 tok/s | ~315 ms ~322 ms ~336 ms | |
#219 | Mixtral-8x7B-v0.1 | 46.7B (7B active) | INT4 | 76 tok/s | ~395 ms | |
#220 | Ministral-8B-2410 | 8B | INT4 Q8 FP16 | 82 tok/s 51 tok/s 28 tok/s | ~358 ms ~365 ms ~381 ms | |
#221 | DeepSeek-R1 1.5B | 1.5B | INT4 Q8 FP16 | 215 tok/s 166 tok/s 110 tok/s | ~80 ms ~81 ms ~84 ms | |
#222 | Llama 3.2 3B | 3B | INT4 Q8 FP16 | 155 tok/s 109 tok/s 65 tok/s | ~145 ms ~148 ms ~154 ms | |
#223 | Phi-3-mini | 3.8B | INT4 Q8 FP16 | 136 tok/s 92 tok/s 54 tok/s | ~179 ms ~183 ms ~190 ms | |
#224 | Gemma 3n E2B IT | 6B (2B active) | INT4 Q8 FP16 | 107 tok/s 92 tok/s 67 tok/s | ~113 ms ~115 ms ~119 ms | |
#225 | Gemma 2 2B | 2B | INT4 Q8 FP16 | 189 tok/s 140 tok/s 89 tok/s | ~103 ms ~105 ms ~109 ms | |
#226 | Mistral-7B-Instruct-v0.1 | 7.3B | INT4 Q8 FP16 | 88 tok/s 55 tok/s 30 tok/s | ~328 ms ~334 ms ~349 ms | |
#227 | Llama 3.2 1B | 1B | INT4 Q8 FP16 | 248 tok/s 202 tok/s 143 tok/s | ~57 ms ~58 ms ~60 ms | |
#228 | Phi-3-medium | 14B | INT4 Q8 FP16 | 52 tok/s 31 tok/s 16 tok/s | ~615 ms ~628 ms ~711 ms | |
#229 | Yi-34B | 34B | INT4 | 19 tok/s | ~1.5s | |
#230 | OLMo 3 7B Instruct | 7B | INT4 Q8 FP16 | 66 tok/s 47 tok/s 28 tok/s | ~319 ms ~325 ms ~340 ms | |
#231 | Qwen2-7B | 7B | INT4 Q8 FP16 | 90 tok/s 57 tok/s 31 tok/s | ~315 ms ~322 ms ~336 ms | |
#232 | Phi-3-small | 7B | INT4 Q8 FP16 | 90 tok/s 57 tok/s 31 tok/s | ~315 ms ~322 ms ~336 ms | |
#233 | Mistral-7B-Instruct-v0.2 | 7.3B | INT4 Q8 FP16 | 88 tok/s 55 tok/s 30 tok/s | ~328 ms ~334 ms ~349 ms | |
#234 | Gemma 3 1B | 1B | INT4 Q8 FP16 | 248 tok/s 202 tok/s 143 tok/s | ~57 ms ~58 ms ~60 ms | |
#236 | Gemma 1 7B | 7B | INT4 Q8 FP16 | 66 tok/s 47 tok/s 28 tok/s | ~319 ms ~325 ms ~340 ms | |
#237 | Gemma 1 2B | 2B | INT4 Q8 FP16 | 200 tok/s 146 tok/s 91 tok/s | ~103 ms ~104 ms ~109 ms | |
#238 | ChatGLM3-6B | 6B | INT4 Q8 FP16 | 73 tok/s 53 tok/s 32 tok/s | ~277 ms ~282 ms ~294 ms | |
#239 | ChatGLM2-6B | 6B | INT4 Q8 FP16 | 73 tok/s 53 tok/s 32 tok/s | ~277 ms ~282 ms ~294 ms | |
#240 | ChatGLM-6B | 6B | INT4 Q8 FP16 | 73 tok/s 53 tok/s 32 tok/s | ~277 ms ~282 ms ~294 ms | |
Unranked Models (Sorted by Speed) | ||||||
- | DINOv3 ViT-L/16 | 0.3B | INT4 Q8 FP16 | 325 tok/s 298 tok/s 252 tok/s | ~22 ms ~22 ms ~22 ms | |
- | Qwen2-0.5B | 0.5B | INT4 Q8 FP16 | 300 tok/s 263 tok/s 208 tok/s | ~31 ms ~32 ms ~33 ms | |
- | Qwen2.5-0.5B | 0.5B | INT4 Q8 FP16 | 300 tok/s 263 tok/s 208 tok/s | ~31 ms ~32 ms ~33 ms | |
- | ERNIE-4.5-0.3B | 0.3B | INT4 Q8 FP16 | 295 tok/s 275 tok/s 235 tok/s | ~22 ms ~22 ms ~23 ms | |
- | ERNIE-4.5-0.3B-Base | 0.3B | INT4 Q8 FP16 | 295 tok/s 275 tok/s 235 tok/s | ~22 ms ~22 ms ~23 ms | |
- | Liquid D1-Omni 600M | 0.6B | INT4 Q8 FP16 | 288 tok/s 249 tok/s 191 tok/s | ~36 ms ~37 ms ~38 ms | |
- | Falcon-1B | 1B | INT4 Q8 FP16 | 259 tok/s 209 tok/s 146 tok/s | ~57 ms ~57 ms ~59 ms | |
- | Falcon3-1B | 1B | INT4 Q8 FP16 | 248 tok/s 202 tok/s 143 tok/s | ~57 ms ~58 ms ~60 ms | |
- | MiniCPM-V 4.7 1B | 1B | INT4 Q8 FP16 | 248 tok/s 202 tok/s 143 tok/s | ~57 ms ~58 ms ~60 ms | |
- | Qwen2-1.5B | 1.5B | INT4 Q8 FP16 | 215 tok/s 166 tok/s 110 tok/s | ~80 ms ~81 ms ~84 ms | |
- | Qwen2.5-1.5B | 1.5B | INT4 Q8 FP16 | 215 tok/s 166 tok/s 110 tok/s | ~80 ms ~81 ms ~84 ms | |
- | Spark X2.5 1.7B | 1.7B | INT4 Q8 FP16 | 203 tok/s 154 tok/s 101 tok/s | ~89 ms ~91 ms ~94 ms | |
- | Hy-MT2 1.8B | 1.8B | INT4 Q8 FP16 | 199 tok/s 150 tok/s 97 tok/s | ~93 ms ~95 ms ~99 ms | |
- | Falcon-3B | 3B | INT4 Q8 FP16 | 163 tok/s 112 tok/s 66 tok/s | ~145 ms ~148 ms ~154 ms | |
- | CroissantLLM Base | 1.3B | INT4 Q8 FP16 | 161 tok/s 138 tok/s 101 tok/s | ~72 ms ~73 ms ~76 ms | |
- | Phi-1 | 1.3B | INT4 Q8 FP16 | 161 tok/s 138 tok/s 101 tok/s | ~72 ms ~73 ms ~76 ms | |
- | Phi-1.5 | 1.3B | INT4 Q8 FP16 | 161 tok/s 138 tok/s 101 tok/s | ~72 ms ~73 ms ~76 ms | |
- | DeepSeek-R1 3B | 3B | INT4 Q8 FP16 | 157 tok/s 109 tok/s 66 tok/s | ~145 ms ~148 ms ~154 ms | |
- | Mellum 2.1 12B A2.5B Thinking | 12B (2.5B active) | INT4 Q8 FP16 | 157 tok/s 116 tok/s 73 tok/s | ~142 ms ~144 ms ~149 ms | |
- | Falcon3-3B | 3B | INT4 Q8 FP16 | 155 tok/s 109 tok/s 65 tok/s | ~145 ms ~148 ms ~154 ms | |
- | LFM2-VL 3B | 3B | INT4 Q8 FP16 | 155 tok/s 109 tok/s 65 tok/s | ~145 ms ~148 ms ~154 ms | |
- | Liquid D1 3B | 3B | INT4 Q8 FP16 | 155 tok/s 109 tok/s 65 tok/s | ~145 ms ~148 ms ~154 ms | |
- | Liquid D1-VL 3B | 3B | INT4 Q8 FP16 | 155 tok/s 109 tok/s 65 tok/s | ~145 ms ~148 ms ~154 ms | |
- | Liquid LFM2-VL 3B D1 | 3B | INT4 Q8 FP16 | 155 tok/s 109 tok/s 65 tok/s | ~145 ms ~148 ms ~154 ms | |
- | Qwen2.5-3B | 3B | INT4 Q8 FP16 | 155 tok/s 109 tok/s 65 tok/s | ~145 ms ~148 ms ~154 ms | |
- | LFM2.5 24B A2B | 24B (2.5B active) | INT4 Q8 | 146 tok/s 105 tok/s | ~160 ms ~175 ms | |
- | ERNIE-4.5-21B-A3B | 21B (3B active) | INT4 Q8 | 137 tok/s 100 tok/s | ~176 ms ~178 ms | |
- | ERNIE-4.5-21B-A3B-Base | 21B (3B active) | INT4 Q8 | 137 tok/s 100 tok/s | ~176 ms ~178 ms | |
- | ERNIE-4.5-VL-28B-A3B | 28B (3B active) | INT4 Q8 | 133 tok/s 93 tok/s | ~184 ms ~202 ms | |
- | ERNIE-4.5-VL-28B-A3B-Base | 28B (3B active) | INT4 Q8 | 133 tok/s 93 tok/s | ~184 ms ~202 ms | |
- | FrogNano 4B | 4B | INT4 Q8 FP16 | 132 tok/s 89 tok/s 51 tok/s | ~188 ms ~191 ms ~199 ms | |
- | Hy-Memory 30B A3B | 30B (3B active) | INT4 Q8 | 132 tok/s 88 tok/s | ~187 ms ~221 ms | |
- | Hy-MT2 30B A3B | 30B (3B active) | INT4 Q8 | 132 tok/s 88 tok/s | ~187 ms ~221 ms | |
- | Spark X2.5 4B | 4B | INT4 Q8 FP16 | 132 tok/s 89 tok/s 51 tok/s | ~188 ms ~191 ms ~199 ms | |
- | MiniCPM-V 4.7 35B A3B | 35B (3B active) | INT4 | 130 tok/s | ~191 ms | |
- | EmbeddingGemma 2 | 2B | INT4 Q8 FP16 | 127 tok/s 105 tok/s 73 tok/s | ~105 ms ~107 ms ~111 ms | |
- | Phi-2 | 2.7B | INT4 Q8 FP16 | 113 tok/s 90 tok/s 60 tok/s | ~135 ms ~137 ms ~143 ms | |
- | MaLLaM-3B | 3B | INT4 Q8 FP16 | 107 tok/s 84 tok/s 56 tok/s | ~148 ms ~150 ms ~157 ms | |
- | SmolLM3 3B | 3B | INT4 Q8 FP16 | 107 tok/s 84 tok/s 56 tok/s | ~148 ms ~150 ms ~157 ms | |
- | Falcon-7B | 7B | INT4 Q8 FP16 | 94 tok/s 58 tok/s 32 tok/s | ~315 ms ~321 ms ~335 ms | |
- | Falcon3-7B | 7B | INT4 Q8 FP16 | 90 tok/s 57 tok/s 31 tok/s | ~315 ms ~322 ms ~336 ms | |
- | Hy-MT2 7B | 7B | INT4 Q8 FP16 | 90 tok/s 57 tok/s 31 tok/s | ~315 ms ~322 ms ~336 ms | |
- | Mistral-7B-v0.1 | 7.3B | INT4 Q8 FP16 | 88 tok/s 55 tok/s 30 tok/s | ~328 ms ~334 ms ~349 ms | |
- | Youtu Parsing Omni | 8B | INT4 Q8 FP16 | 84 tok/s 52 tok/s 28 tok/s | ~357 ms ~365 ms ~381 ms | |
- | LensVLM 9B | 9B | INT4 Q8 FP16 | 75 tok/s 46 tok/s 25 tok/s | ~401 ms ~409 ms ~428 ms | |
- | MiMo V2.6 Distill Qwen 9B | 9B | INT4 Q8 FP16 | 75 tok/s 46 tok/s 25 tok/s | ~401 ms ~409 ms ~428 ms | |
- | ChatGLM3-6B-32K | 6B | INT4 Q8 FP16 | 73 tok/s 53 tok/s 32 tok/s | ~277 ms ~282 ms ~294 ms | |
- | Yi-6B | 6B | INT4 Q8 FP16 | 73 tok/s 53 tok/s 32 tok/s | ~277 ms ~282 ms ~294 ms | |
- | Falcon3-10B | 10B | INT4 Q8 FP16 | 69 tok/s 42 tok/s 23 tok/s | ~443 ms ~453 ms ~473 ms | |
- | Kimi-VL-A3B-Instruct | 16B (3B active) | INT4 Q8 | 69 tok/s 60 tok/s | ~176 ms ~178 ms | |
- | Falcon2-11B | 11B | INT4 Q8 FP16 | 66 tok/s 39 tok/s 21 tok/s | ~485 ms ~496 ms ~518 ms | |
- | Hunyuan Lite | 7B | INT4 Q8 FP16 | 66 tok/s 47 tok/s 28 tok/s | ~319 ms ~325 ms ~340 ms | |
- | OLMo 3 7B Base | 7B | INT4 Q8 FP16 | 66 tok/s 47 tok/s 28 tok/s | ~319 ms ~325 ms ~340 ms | |
- | SEA-LION-7B | 7.1B | INT4 Q8 FP16 | 65 tok/s 46 tok/s 28 tok/s | ~324 ms ~330 ms ~344 ms | |
- | SEA-LION-7B-Instruct | 7.1B | INT4 Q8 FP16 | 65 tok/s 46 tok/s 28 tok/s | ~324 ms ~330 ms ~344 ms | |
- | Sahabat-AI-Llama3-8B-Instruct | 8B | INT4 Q8 FP16 | 60 tok/s 42 tok/s 25 tok/s | ~362 ms ~369 ms ~385 ms | |
- | Typhoon-2-8B | 8B | INT4 Q8 FP16 | 60 tok/s 42 tok/s 25 tok/s | ~362 ms ~369 ms ~385 ms | |
- | GLM-4-9B | 9B | INT4 Q8 FP16 | 55 tok/s 38 tok/s 22 tok/s | ~406 ms ~414 ms ~432 ms | |
- | GLM-4-9B-Chat | 9B | INT4 Q8 FP16 | 55 tok/s 38 tok/s 22 tok/s | ~406 ms ~414 ms ~432 ms | |
- | GLM-4-9B-Chat-1M | 9B | INT4 Q8 FP16 | 55 tok/s 38 tok/s 22 tok/s | ~406 ms ~414 ms ~432 ms | |
- | Yi-9B | 9B | INT4 Q8 FP16 | 55 tok/s 38 tok/s 22 tok/s | ~406 ms ~414 ms ~432 ms | |
- | Sahabat-AI-Gemma2-9B | 9.2B | INT4 Q8 FP16 | 54 tok/s 37 tok/s 22 tok/s | ~414 ms ~422 ms ~441 ms | |
- | Sahabat-AI-Gemma2-9B-Instruct | 9.2B | INT4 Q8 FP16 | 54 tok/s 37 tok/s 22 tok/s | ~414 ms ~422 ms ~441 ms | |
- | GreenMind-14B-R1 | 14B | INT4 Q8 FP16 | 38 tok/s 26 tok/s 13 tok/s | ~622 ms ~634 ms ~773 ms | |
- | AREX-2 | 27B | INT4 Q8 | 29 tok/s 16 tok/s | ~1.2s ~1.3s | |
- | PPLX Decider v1 27B | 27B | INT4 Q8 | 29 tok/s 16 tok/s | ~1.2s ~1.3s | |
- | PPLX Decider v1.1 27B | 27B | INT4 Q8 | 29 tok/s 16 tok/s | ~1.2s ~1.3s | |
- | Agnes 3.0 Flash | 33B | INT4 | 24 tok/s | ~1.4s | |
- | Falcon-40B | 40B | INT4 | 21 tok/s | ~1.7s | |
Inference Capacity
Maximum model parameter class runnable by weight precision.
Context:
4-bit
32B Models
8-bit
14B Models
16-bit
14B Models
Fine-Tuning Capacity
Hardware capacity for local training and adapter fine-tuning.
Context Length:
QLoRA
32B Models
LoRA
14B Models
Full Parameter Training
3B Models
Workload Recommendations
Guidance for optimal precision and operational ceilings on NVIDIA RTX 5000 Ada Generation (32GB).
Inference Sweet Spot
Optimized for 14B models at uncompressed or Q8 precision, and 32B models at Q4 with up to 32k 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