Mac Pro M2 Ultra 192GB

macOS Ultra tier · Apple M2 Ultra · 192 GB RAM · 800 GB/s bandwidth
View device page →
BENCHMARK RESULTS (76)
Model Quant Measured Estimated RAM used App Source Date
StableLM 2 1.6B StableLM 2 1.6B FP16 223 tok/s 200 tok/s +12% 3.8 GB Ollama community 2026-04-10
🔢 all-MiniLM-L6-v2 FP16 222 tok/s 200 tok/s +11% 0.1 GB Ollama community 2026-03-03
LFM2.5 1.2B LFM2.5 1.2B FP16 221 tok/s 200 tok/s +11% 2.5 GB Ollama community 2026-05-11
SmolLM2 1.7B SmolLM2 1.7B FP16 221 tok/s 200 tok/s +11% 3.8 GB Xybrid CLI editorial 2026-01-03
Whisper Medium Whisper Medium FP16 220 tok/s 200 tok/s +10% 1.8 GB Xybrid CLI community 2026-03-13
Ternary Bonsai 1.7B Ternary Bonsai 1.7B Q2 219 tok/s 200 tok/s +10% 0.4 GB Xybrid CLI editorial 2026-01-02
🎙️ Distil-Whisper Large V3 FP16 219 tok/s 200 tok/s +10% 1.8 GB Xybrid CLI community 2026-04-19
👁️ SmolVLM 500M FP16 217 tok/s 200 tok/s +9% 1.2 GB Xybrid CLI community 2026-01-03
🗣️ NeuTTS Air FP16 217 tok/s 200 tok/s +9% 1.8 GB Xybrid CLI community 2026-03-15
🔢 GTE Large FP16 216 tok/s 200 tok/s +8% 0.7 GB Ollama community 2026-03-28
SmolLM2 135M SmolLM2 135M FP16 214 tok/s 200 tok/s +7% 0.3 GB LM Studio community 2026-05-10
🎨 Stable Diffusion Turbo FP16 212 tok/s 200 tok/s +6% 2.7 GB LM Studio editorial 2026-02-25
Qwen 3.5 0.8B Qwen 3.5 0.8B FP16 210 tok/s 200 tok/s +5% 2.3 GB Ollama editorial 2026-02-27
Gemma 3 1B Gemma 3 1B FP16 207 tok/s 200 tok/s +4% 2.8 GB LM Studio editorial 2026-02-25
🔢 BGE Small FP16 207 tok/s 200 tok/s +4% 0.1 GB Jan community 2026-02-06
🔢 Nomic Embed Text FP16 207 tok/s 200 tok/s +4% 0.3 GB LM Studio community 2026-03-28
Ternary Bonsai 4B Ternary Bonsai 4B Q2 206 tok/s 200 tok/s +3% 1.0 GB LM Studio editorial 2026-05-11
Whisper Large V3 Whisper Large V3 FP16 205 tok/s 200 tok/s +3% 3.4 GB Xybrid CLI community 2026-02-08
Qwen 2.5 Coder 1.5B Qwen 2.5 Coder 1.5B FP16 203 tok/s 200 tok/s +2% 3.6 GB LM Studio editorial 2026-05-10
Llama 3.2 1B Llama 3.2 1B FP16 202 tok/s 200 tok/s +1% 3.1 GB Jan community 2026-03-12
Whisper Small Whisper Small FP16 202 tok/s 200 tok/s +1% 0.5 GB MacWhisper community 2026-01-03
🗣️ Dia 1.6B FP16 200 tok/s 200 tok/s +0% 3.7 GB Piper community 2026-02-08
👁️ Moondream 2B FP16 196 tok/s 200 tok/s -2% 4.2 GB Xybrid CLI editorial 2026-04-05
SmolLM2 360M SmolLM2 360M FP16 195 tok/s 200 tok/s -2% 1.1 GB Ollama community 2026-04-21
Ternary Bonsai 8B Ternary Bonsai 8B Q2 195 tok/s 200 tok/s -2% 1.7 GB Xybrid CLI community 2026-03-27
🗣️ Kokoro 82M FP16 194 tok/s 200 tok/s -3% 0.4 GB Xybrid CLI community 2026-03-26
Whisper Tiny Whisper Tiny FP16 193 tok/s 200 tok/s -4% 0.1 GB Xybrid CLI community 2026-04-04
Qwen 2.5 Coder 0.5B Qwen 2.5 Coder 0.5B FP16 191 tok/s 200 tok/s -4% 1.3 GB Ollama community 2026-01-06
🗣️ OuteTTS 0.3 500M FP16 191 tok/s 200 tok/s -4% 1.1 GB Xybrid CLI community 2026-04-06
🗣️ KittenTTS Nano FP16 190 tok/s 200 tok/s -5% 0.1 GB Piper community 2026-01-18
Gemma 4 E2B Gemma 4 E2B FP16 187 tok/s 178 tok/s +5% 5.1 GB LM Studio editorial 2026-04-11
Mistral 7B Mistral 7B Q4 186 tok/s 54 tok/s +244% 4.3 GB LM Studio editorial 2026-03-20
TinyLlama 1.1B TinyLlama 1.1B FP16 185 tok/s 200 tok/s -7% 2.6 GB Jan community 2026-04-10
Bonsai Image 4B Bonsai Image 4B Q2 185 tok/s 200 tok/s -7% 2.1 GB Ollama community 2026-04-18
Wav2Vec2 Base Wav2Vec2 Base FP16 185 tok/s 200 tok/s -7% 0.3 GB Xybrid CLI community 2026-01-18
Bonsai 8B (1-bit) Bonsai 8B (1-bit) Q1 182 tok/s 200 tok/s -9% 1.4 GB LM Studio community 2026-02-11
Qwen 2.5 0.5B Qwen 2.5 0.5B FP16 181 tok/s 200 tok/s -9% 1.4 GB Jan community 2026-05-16
Qwen 3.5 2B Qwen 3.5 2B FP16 181 tok/s 190 tok/s -5% 5.4 GB Ollama editorial 2026-01-05
🧠 DeepSeek R1 Distill 1.5B FP16 180 tok/s 200 tok/s -10% 3.6 GB Xybrid CLI community 2026-02-25
🗣️ KittenTTS Mini FP16 177 tok/s 200 tok/s -11% 0.2 GB Xybrid CLI community 2026-05-14
Phi-4 Mini Phi-4 Mini Q8 174 tok/s 103 tok/s +69% 4.6 GB Ollama editorial 2026-03-20
🎨 Stable Diffusion 3.5 Medium FP16 155 tok/s 174 tok/s -11% 5.3 GB LM Studio editorial 2026-01-06
Llama 3.2 3B Llama 3.2 3B FP16 133 tok/s 121 tok/s +10% 7.8 GB Xybrid CLI community 2026-04-19
🎨 SDXL Turbo FP16 133 tok/s 123 tok/s +8% 7.2 GB Ollama community 2026-04-06
👨‍💻 StarCoder2 3B FP16 132 tok/s 129 tok/s +2% 6.9 GB LM Studio editorial 2026-04-05
Qwen 2.5 3B Qwen 2.5 3B FP16 121 tok/s 125 tok/s -3% 7.6 GB Xybrid CLI community 2026-05-14
Qwen 2.5 Coder 3B Qwen 2.5 Coder 3B FP16 117 tok/s 125 tok/s -6% 8.1 GB LM Studio editorial 2026-05-23
Phi-4 Mini Phi-4 Mini FP16 109 tok/s 103 tok/s +6% 8.9 GB Ollama editorial 2026-01-17
Gemma 4 E4B Gemma 4 E4B FP16 99 tok/s 89 tok/s +11% 11.2 GB Ollama community 2026-02-21
Gemma 3 4B Gemma 3 4B FP16 97 tok/s 90 tok/s +8% 9.8 GB Xybrid CLI community 2026-01-02
Gemma 3n E2B Gemma 3n E2B FP16 95 tok/s 90 tok/s +6% 10.2 GB Jan community 2026-03-11
Qwen 3.5 4B Qwen 3.5 4B FP16 87 tok/s 98 tok/s -11% 10.3 GB Xybrid CLI editorial 2026-04-17
👨‍💻 DeepSeek Coder 6.7B FP16 64 tok/s 58 tok/s +10% 14.8 GB Ollama community 2026-04-08
Gemma 3n E4B Gemma 3n E4B FP16 62 tok/s 58 tok/s +7% 16.4 GB Jan community 2026-03-16
Qwen 2.5 Coder 7B Qwen 2.5 Coder 7B FP16 58 tok/s 52 tok/s +12% 18.6 GB Xybrid CLI community 2026-04-21
🧠 DeepSeek R1 Distill 7B FP16 56 tok/s 52 tok/s +8% 18.2 GB Jan community 2026-03-12
🧠 DeepSeek R1 Distill 8B FP16 56 tok/s 50 tok/s +12% 18.3 GB LM Studio editorial 2026-01-06
👁️ LLaVA 1.6 7B FP16 56 tok/s 56 tok/s +0% 15.4 GB Ollama community 2026-01-04
Qwen 2.5 7B Qwen 2.5 7B FP16 55 tok/s 52 tok/s +6% 19.9 GB LM Studio editorial 2026-03-03
LFM2.5 8B A1B LFM2.5 8B A1B FP16 53 tok/s 48 tok/s +10% 20.5 GB Xybrid CLI community 2026-04-18
Qwen 2.5 VL 7B Qwen 2.5 VL 7B FP16 53 tok/s 49 tok/s +8% 17.8 GB Xybrid CLI community 2026-01-05
🧠 Qwen3 8B FP16 52 tok/s 48 tok/s +8% 18.3 GB Xybrid CLI community 2026-02-08
Llama 3.1 8B Llama 3.1 8B FP16 51 tok/s 50 tok/s +2% 17.7 GB Ollama community 2026-02-10
Mistral 7B Mistral 7B FP16 51 tok/s 54 tok/s -6% 17.8 GB Jan community 2026-05-12
Qwen 3.5 9B Qwen 3.5 9B FP16 46 tok/s 43 tok/s +7% 21.5 GB Ollama community 2026-05-11
Gemma 3 12B Gemma 3 12B FP16 36 tok/s 32 tok/s +13% 27.1 GB Jan community 2026-04-06
Mistral Nemo 12B Mistral Nemo 12B FP16 36 tok/s 33 tok/s +9% 28.9 GB Xybrid CLI community 2026-04-22
🎨 FLUX.1 Schnell FP16 35 tok/s 34 tok/s +3% 26.3 GB Ollama community 2026-01-05
Phi-4 Medium Phi-4 Medium FP16 26 tok/s 29 tok/s -10% 31.6 GB LM Studio editorial 2026-02-10
Gemma 4 26B A4B Gemma 4 26B A4B FP16 14 tok/s 15 tok/s -7% 64.6 GB Xybrid CLI community 2026-01-19
Gemma 4 31B Gemma 4 31B FP16 14 tok/s 13 tok/s +8% 69.0 GB Ollama community 2026-04-10
👨‍💻 Laguna XS.2 FP16 11 tok/s 12 tok/s -8% 75.6 GB Ollama community 2026-03-05
Qwen 3.5 35B A3B Qwen 3.5 35B A3B FP16 10 tok/s 11 tok/s -9% 78.4 GB LM Studio editorial 2026-03-12
Qwen3 Coder Next Qwen3 Coder Next Q8 9 tok/s 9 tok/s +0% 96.5 GB Ollama community 2026-01-03
Llama 4 Scout Llama 4 Scout Q8 7 tok/s 7 tok/s +0% 125.9 GB Ollama community 2026-01-02
DeepSeek V4 Flash DeepSeek V4 Flash Q3 7 tok/s 7 tok/s +0% 152.4 GB LM Studio editorial 2026-01-01
← All benchmarks