Apple M5 Pro (48GB)
The Apple M5 Pro (48GB) has 48 GB VRAM and 307 GB/s memory bandwidth. It can run 49 of our 81 tracked models natively in VRAM at 8k context.
With 48 GB LPDDR5X, the Apple M5 Pro (48GB) is a laptop-tier GPU that can run 49 models natively. It handles 70B-class models at Q4 quantization.
The Apple M5 Pro (48GB) is purpose-built for local LLM inference on MacBook Pro and Mac Studio, offering 48GB of unified memory and 307 GB/s bandwidth. It runs Qwen 3.6 35B at Q8_0 and Gemma 4 31B at Q8_0 or F16 entirely in-memory without CPU offload, covering both of today's highest-search-volume open-weight models via MLX and llama.cpp. The 15-core M5 Pro CPU and Apple Silicon's power efficiency make this the go-to on-device AI configuration for professionals who want top-tier models in a portable Mac form factor.
Apple M5 Pro (48GB): 2026 top-tier M5 Pro configuration — 48GB unified LPDDR5X at 307 GB/s, 15-core CPU (5P+10E).
Qwen 3.6 35B and Gemma 4 31B both fit at Q8_0 or F16 entirely in-memory, covering the two highest-search-volume open-weight models via MLX and llama.cpp. ~10-16 t/s for 7B via MLX.
MLX and llama.cpp Metal fully supported. Delivers Max-class model compatibility at Pro-tier bandwidth and power efficiency in a portable form factor.
| Vendor | Apple |
| Architecture | Apple M5 Pro |
| CPU cores | 15 (5S + 10P) |
| VRAM | 48 GB (unified) |
| Memory type | LPDDR5X |
| Memory bandwidth | 307 GB/s |
| Compute backend | METAL |
| Tier | Laptop |
| Released | 2026 |
| Models (native) | 49 / 81 |
| Models (offload) | 0 / 81 |
Popular models for this GPU
Models this GPU runs natively in VRAM (49)
- Qwen 2.5 72B Instruct72B · MMLU-Pro 71.1Q2_K · ~8.2 t/s
- Llama 3.3 70B Instruct70B · MMLU-Pro 68.9Q2_K · ~8.4 t/s
- DeepSeek R1 Distill Llama 70B70B · MMLU-Pro 70.0Q2_K · ~8.4 t/s
- Llama 3.1 70B Instruct70B · MMLU-Pro 66.4Q2_K · ~8.4 t/s
- Mixtral 8x7B Instruct v0.146.7B · MMLU-Pro 29.7Q5_K_M · ~7.8 t/s
Show 44 more
- Command-R 35B35B · MMLU-Pro 33.0Q5_K_M · ~6.9 t/s
- Qwen 3.5 35B-A3B (MoE)35B · MMLU-Pro 85.3Q6_K · ~29.3 t/s
- Qwen 3.6 35B35B · MMLU-Pro 85.2Q6_K · ~8 t/s
- Yi 1.5 34B Chat34.4B · MMLU-Pro 37.0Q6_K · ~8.1 t/s
- Qwen3 32B32.8B · MMLU-Pro 65.5Q6_K · ~8.7 t/s
- Qwen 2.5 32B Instruct32.5B · MMLU-Pro 69.0Q6_K · ~8.5 t/s
- Qwen 2.5 Coder 32B Instruct32.5B · MMLU-Pro 50.4Q6_K · ~8.5 t/s
- DeepSeek R1 Distill Qwen 32B32.5B · MMLU-Pro 65.0Q6_K · ~8.5 t/s
- Nemotron 3 Nano 30B32B · MMLU-Pro 78.3Q8_0 · ~22.2 t/s
- Gemma 4 31B31B · MMLU-Pro 85.2Q6_K · ~8.6 t/s
- Qwen3 30B-A3B (MoE)30B · MMLU-Pro 61.5Q8_0 · ~21.5 t/s
- Gemma 2 27B Instruct27.2B · MMLU-Pro 38.0Q8_0 · ~7.7 t/s
- Gemma 3 27B Instruct27B · MMLU-Pro 67.5Q8_0 · ~8.1 t/s
- Qwen 3.6 27B27B · MMLU-Pro 86.2Q8_0 · ~8.4 t/s
- Bonsai 27B27B · MMLU-Pro 81.5Ternary (Q2_0) · ~27.9 t/s
- Gemma 4 26B (MoE)26B · MMLU-Pro 82.6Q8_0 · ~17.7 t/s
- Mistral Small 3.1 24B Instruct24B · MMLU-Pro 66.8Q8_0 · ~9.1 t/s
- Mistral Small 22B22.2B · MMLU-Pro 49.2Q8_0 · ~9.6 t/s
- GPT-OSS 20B21B · MMLU-Pro 67.9Q8_0 · ~16.8 t/s
- Qwen3 14B14.8B · MMLU-Pro 61.0BF16 · ~7.9 t/s
- Qwen 2.5 14B Instruct14.7B · MMLU-Pro 63.7BF16 · ~7.9 t/s
- Phi-4 14B Instruct14B · MMLU-Pro 70.4BF16 · ~8.4 t/s
- Mistral Nemo 12B Instruct12.2B · MMLU-Pro 35.6BF16 · ~9.5 t/s
- Gemma 3 12B Instruct12.2B · MMLU-Pro 60.6BF16 · ~9.6 t/s
- Gemma 4 12B (Unified)12B · MMLU-Pro 77.2BF16 · ~9 t/s
- Gemma 2 9B Instruct9.2B · MMLU-Pro 32.0BF16 · ~11.6 t/s
- Llama 3.1 8B Instruct8B · MMLU-Pro 48.3FP32 · ~7.4 t/s
- DeepSeek R1 Distill Llama 8B8B · MMLU-Pro 41.0FP32 · ~7.4 t/s
- Qwen3 8B8B · MMLU-Pro 56.7FP32 · ~7.4 t/s
- Qwen 2.5 7B Instruct7.6B · MMLU-Pro 56.3FP32 · ~8 t/s
- Mistral 7B Instruct v0.37.25B · MMLU-Pro 30.0FP32 · ~8.2 t/s
- Gemma 3 4B Instruct4B · MMLU-Pro 43.6FP32 · ~14.9 t/s
- Gemma 4 E4B4B · MMLU-Pro 69.4FP32 · ~14.4 t/s
- Phi-3.5 Mini Instruct3.8B · MMLU-Pro 47.4FP32 · ~13.3 t/s
- Llama 3.2 3B Instruct3.2B · MMLU-Pro 24.0FP32 · ~17.9 t/s
- Qwen 2.5 3B Instruct3.1B · MMLU-Pro 32.4FP32 · ~19.3 t/s
- Gemma 2 2B Instruct2.6B · MMLU-Pro 17.8FP32 · ~21.8 t/s
- Gemma 4 E2B2B · MMLU-Pro 60.0FP32 · ~29.2 t/s
- SmolLM2 1.7B Instruct1.7B · MMLU-Pro 19.0FP32 · ~29.2 t/s
- Qwen 2.5 1.5B Instruct1.5B · MMLU-Pro 16.8FP32 · ~39.4 t/s
- Llama 3.2 1B Instruct1.24B · MMLU-Pro 12.5FP32 · ~47 t/s
- Gemma 3 1B Instruct1B · MMLU-Pro 14.7FP32 · ~56.8 t/s
- Qwen 2.5 0.5B Instruct0.5B · MMLU-Pro 10.0FP32 · ~116.9 t/s
- SmolLM2 360M Instruct0.36B · MMLU-Pro 8.0FP32 · ~138.3 t/s
Too large for this GPU (32)
- Mixtral 8x22B Instruct v0.1
- Llama 3.1 405B Instruct
- DeepSeek V3 671B
- DeepSeek R1 671B
- Llama 4 Scout 109B
- Llama 4 Maverick 400B
- Qwen3 235B-A22B (MoE)
- MiniMax M1 456B
- GPT-OSS 120B
- GLM-4.5 355B
- GLM-4.5 Air 106B
- GLM-4.6 355B
- GLM-4.6V 106B
- GLM-4.7 358B
- Qwen 3.5 122B-A10B (MoE)
- MiniMax M2.5 229B
- GLM-5 744B
- MiniMax M2.7 229B
- Nemotron 3 Super 120B
- Kimi K2.6
- GLM-5.1 754B
- DeepSeek V4 Pro 1.6T
- DeepSeek V4 Flash 284B
- GLM-5.2 753B
- Nemotron 3 Ultra 550B-A55B
- Step 3.5 Flash
- Step 3.7 Flash
- MiMo V2.5 Pro
- Kimi K2.5
- MiniMax M3
- Inkling
- Kimi K3
Frequently asked questions
- How much VRAM does the Apple M5 Pro (48GB) have?
- The Apple M5 Pro (48GB) has 48 GB of LPDDR5X with 307 GB/s memory bandwidth (unified system memory, shared between CPU and GPU).
- What is the Apple M5 Pro (48GB) best for?
- With 48 GB of VRAM, the Apple M5 Pro (48GB) is ideal for running 70B-class models at Q4 quantization and large MoE models — a workstation sweet spot for local inference.
- What LLMs can the Apple M5 Pro (48GB) run locally?
- The Apple M5 Pro (48GB) can run 49 of the 81 open-weight models tracked by CanItRun natively in VRAM at 8k context. Top options include: Llama 3.3 70B Instruct at Q2_K, Llama 3.1 8B Instruct at FP32, Llama 3.2 3B Instruct at FP32.
- Can the Apple M5 Pro (48GB) run Llama 3.3 70B Instruct?
- Yes. The Apple M5 Pro (48GB) runs Llama 3.3 70B Instruct natively in VRAM at Q2_K quantization, achieving approximately 8.4 tokens per second.
- Can the Apple M5 Pro (48GB) run Qwen 3.6 27B?
- Yes. The Apple M5 Pro (48GB) runs Qwen 3.6 27B natively in VRAM at Q8_0 quantization, achieving approximately 8.4 tokens per second.
- Can the Apple M5 Pro (48GB) run Llama 3.1 8B Instruct?
- Yes. The Apple M5 Pro (48GB) runs Llama 3.1 8B Instruct natively in VRAM at FP32 quantization, achieving approximately 7.4 tokens per second.