Apple M5 Max (48GB)
The Apple M5 Max (48GB) has 48 GB VRAM and 614 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 Max (48GB) is a laptop-tier GPU that can run 49 models natively. It handles 70B-class models at Q4 quantization.
The Apple M5 Max (48GB) offers 48GB of unified memory and 614 GB/s bandwidth, striking a practical balance between cost and on-device AI capability on Mac. Qwen 3.6 35B runs at Q4_K_M with comfortable headroom, and Gemma 4 31B fits at Q4_K_M or Q8_0 — both fully in-memory without CPU offload. Powered by the 18-core M5 Max and optimized for MLX and llama.cpp, it's the sweet spot for Apple Silicon users who want today's best open-weight models without stepping up to the maximum memory tier.
Apple M5 Max (48GB): 2026 entry M5 Max — 48GB unified LPDDR5X at 614 GB/s, 18-core CPU (6P+12E).
Qwen 3.6 35B fits at Q4_K_M with comfortable headroom; Gemma 4 31B fits at Q4_K_M or Q8_0, both fully in-memory. ~15-24 t/s for 7B via MLX.
MLX and llama.cpp Metal fully optimized. The practical balance point between cost and on-device capability in the M5 Max line.
| Vendor | Apple |
| Architecture | Apple M5 Max |
| CPU cores | 18 (6S + 12P) |
| VRAM | 48 GB (unified) |
| Memory type | LPDDR5X |
| Memory bandwidth | 614 GB/s |
| Compute backend | METAL |
| Tier | Laptop |
| Released | 2026 |
| Models (native) | 49 / 81 |
| Models (offload) | 0 / 81 |
Software: MLX gives the best performance on Apple Silicon; llama.cpp Metal backend is a solid alternative. Both are well-supported by Ollama.
Popular models for this GPU
Models this GPU runs natively in VRAM (49)
- Qwen 2.5 72B Instruct72B · MMLU-Pro 71.1Q2_K · ~16.3 t/s
- Llama 3.3 70B Instruct70B · MMLU-Pro 68.9Q2_K · ~16.7 t/s
- DeepSeek R1 Distill Llama 70B70B · MMLU-Pro 70.0Q2_K · ~16.7 t/s
- Llama 3.1 70B Instruct70B · MMLU-Pro 66.4Q2_K · ~16.7 t/s
- Mixtral 8x7B Instruct v0.146.7B · MMLU-Pro 29.7Q5_K_M · ~15.5 t/s
- Command-R 35B35B · MMLU-Pro 33.0Q5_K_M · ~13.8 t/s
- Qwen 3.5 35B-A3B (MoE)35B · MMLU-Pro 84.2Q6_K · ~54.5 t/s
- Qwen 3.6 35B35B · MMLU-Pro 85.2Q6_K · ~15.9 t/s
- Yi 1.5 34B Chat34.4B · MMLU-Pro 37.0Q6_K · ~16.2 t/s
- Qwen3 32B32.8B · MMLU-Pro 65.5Q6_K · ~17.4 t/s
- Qwen 2.5 32B Instruct32.5B · MMLU-Pro 69.0Q6_K · ~17 t/s
- Qwen 2.5 Coder 32B Instruct32.5B · MMLU-Pro 50.4Q6_K · ~17 t/s
- DeepSeek R1 Distill Qwen 32B32.5B · MMLU-Pro 65.0Q6_K · ~17 t/s
- Nemotron 3 Nano 30B32B · MMLU-Pro 78.3Q8_0 · ~44.4 t/s
- Gemma 4 31B31B · MMLU-Pro 85.2Q6_K · ~17.1 t/s
- Qwen3 30B-A3B (MoE)30B · MMLU-Pro 61.5Q8_0 · ~43 t/s
- Gemma 2 27B Instruct27.2B · MMLU-Pro 38.0Q8_0 · ~15.3 t/s
- Gemma 3 27B Instruct27B · MMLU-Pro 67.5Q8_0 · ~16.2 t/s
- Qwen 3.6 27B27B · MMLU-Pro 86.2Q8_0 · ~16.8 t/s
- Bonsai 27B27B · MMLU-Pro 81.5Ternary (Q2_0) · ~55.8 t/s
- Gemma 4 26B (MoE)26B · MMLU-Pro 82.6Q8_0 · ~33 t/s
- Mistral Small 3.1 24B Instruct24B · MMLU-Pro 66.8Q8_0 · ~18.3 t/s
- Mistral Small 22B22.2B · MMLU-Pro 49.2Q8_0 · ~19.3 t/s
- GPT-OSS 20B21B · MMLU-Pro 67.9Q8_0 · ~33.7 t/s
- Qwen3 14B14.8B · MMLU-Pro 61.0BF16 · ~15.9 t/s
- Qwen 2.5 14B Instruct14.7B · MMLU-Pro 63.7BF16 · ~15.8 t/s
- Phi-4 14B Instruct14B · MMLU-Pro 70.4BF16 · ~16.7 t/s
- Mistral Nemo 12B Instruct12.2B · MMLU-Pro 35.6BF16 · ~19.1 t/s
- Gemma 3 12B Instruct12.2B · MMLU-Pro 60.6BF16 · ~19.3 t/s
- Gemma 4 12B (Unified)12B · MMLU-Pro 77.2BF16 · ~18 t/s
- Gemma 2 9B Instruct9.2B · MMLU-Pro 32.0BF16 · ~23.1 t/s
- Llama 3.1 8B Instruct8B · MMLU-Pro 48.3FP32 · ~14.9 t/s
- DeepSeek R1 Distill Llama 8B8B · MMLU-Pro 41.0FP32 · ~14.9 t/s
- Qwen3 8B8B · MMLU-Pro 56.7FP32 · ~14.8 t/s
- Qwen 2.5 7B Instruct7.6B · MMLU-Pro 56.3FP32 · ~15.9 t/s
- Mistral 7B Instruct v0.37.25B · MMLU-Pro 30.0FP32 · ~16.3 t/s
- Gemma 3 4B Instruct4B · MMLU-Pro 43.6FP32 · ~29.8 t/s
- Gemma 4 E4B4B · MMLU-Pro 69.4FP32 · ~28.9 t/s
- Phi-3.5 Mini Instruct3.8B · MMLU-Pro 47.4FP32 · ~26.7 t/s
- Llama 3.2 3B Instruct3.2B · MMLU-Pro 24.0FP32 · ~35.8 t/s
- Qwen 2.5 3B Instruct3.1B · MMLU-Pro 32.4FP32 · ~38.7 t/s
- Gemma 2 2B Instruct2.6B · MMLU-Pro 17.8FP32 · ~43.6 t/s
- Gemma 4 E2B2B · MMLU-Pro 60.0FP32 · ~58.5 t/s
- SmolLM2 1.7B Instruct1.7B · MMLU-Pro 19.0FP32 · ~58.4 t/s
- Qwen 2.5 1.5B Instruct1.5B · MMLU-Pro 16.8FP32 · ~78.8 t/s
- Llama 3.2 1B Instruct1.24B · MMLU-Pro 12.5FP32 · ~93.9 t/s
- Gemma 3 1B Instruct1B · MMLU-Pro 14.7FP32 · ~113.5 t/s
- Qwen 2.5 0.5B Instruct0.5B · MMLU-Pro 10.0FP32 · ~233.8 t/s
- SmolLM2 360M Instruct0.36B · MMLU-Pro 8.0FP32 · ~276.6 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 Max (48GB) have?
- The Apple M5 Max (48GB) has 48 GB of LPDDR5X with 614 GB/s memory bandwidth (unified system memory, shared between CPU and GPU).
- What is the Apple M5 Max (48GB) best for?
- With 48 GB of VRAM, the Apple M5 Max (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 Max (48GB) run locally?
- The Apple M5 Max (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 Max (48GB) run Llama 3.3 70B Instruct?
- Yes. The Apple M5 Max (48GB) runs Llama 3.3 70B Instruct natively in VRAM at Q2_K quantization, achieving approximately 16.7 tokens per second.
- Can the Apple M5 Max (48GB) run Qwen 3.6 27B?
- Yes. The Apple M5 Max (48GB) runs Qwen 3.6 27B natively in VRAM at Q8_0 quantization, achieving approximately 16.8 tokens per second.
- Can the Apple M5 Max (48GB) run Llama 3.1 8B Instruct?
- Yes. The Apple M5 Max (48GB) runs Llama 3.1 8B Instruct natively in VRAM at FP32 quantization, achieving approximately 14.9 tokens per second.