Apple · M4 Pro · Mini PC
Best AI Models for Mac Mini M4 Pro (48 GB)
Runs 14B and 35B MoE models well; dense 27–31B fits but is slow at 11 tok/s, and 70B only just fits in 44.5 GB.
| # | Model | Speed | llmrun Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 43.2 t/s | 43.2 tok/s | Q4_K_M · 22.0 GB | |
| 2 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 49.1 t/s | 49.1 tok/s | Q4_K_M · 16.5 GB | |
| 3 | Qwen3 30B A3B Thinking 2507Alibaba Q4_K_M · 19.4 GB · 49.1 t/s | 49.1 tok/s | Q4_K_M · 19.4 GB | |
| 4 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 54.5 t/s | 54.5 tok/s | Q4_K_M · 13.1 GB | |
| 5 | Q4_K_M · 19.3 GB · 55.7 t/s | 55.7 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Coding Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 43.2 t/s | 43.2 tok/s | Q4_K_M · 22.0 GB | |
| 2 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 49.1 t/s | 49.1 tok/s | Q4_K_M · 16.5 GB | |
| 3 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 54.5 t/s | 54.5 tok/s | Q4_K_M · 13.1 GB | |
| 4 | Q4_K_M · 7.9 GB · 36.1 t/s | 36.1 tok/s | Q4_K_M · 7.9 GB | |
| 5 | Q4_K_M · 19.3 GB · 55.7 t/s | 55.7 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Reasoning Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 43.2 t/s | 43.2 tok/s | Q4_K_M · 22.0 GB | |
| 2 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 49.1 t/s | 49.1 tok/s | Q4_K_M · 16.5 GB | |
| 3 | Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 30.0 t/s | 30.0 tok/s | Q4_K_M · 6.6 GB | |
| 4 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 54.5 t/s | 54.5 tok/s | Q4_K_M · 13.1 GB | |
| 5 | Q4_K_M · 19.3 GB · 55.7 t/s | 55.7 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Math Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 43.2 t/s | 43.2 tok/s | Q4_K_M · 22.0 GB | |
| 2 | Qwen3.8 27BAlibaba Q4_K_M · 17.9 GB · 11.0 t/s | 11.0 tok/s | Q4_K_M · 17.9 GB | |
| 3 | Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 11.0 t/s | 11.0 tok/s | Q4_K_M · 17.9 GB | |
| 4 | Q4_K_M · 19.3 GB · 55.7 t/s | 55.7 tok/s | Q4_K_M · 19.3 GB | |
| 5 | GLM 4.7 FlashZ.ai Q4_K_M · 19.9 GB · 40.5 t/s | 40.5 tok/s | Q4_K_M · 19.9 GB |
| # | Model | Speed ↓ | llmrun Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.5 4BAlibabaHow to run → Q4_K_M · 3.4 GB · 58.6 t/s | 58.6 tok/s | Q4_K_M · 3.4 GB | |
| 2 | Q4_K_M · 19.3 GB · 55.7 t/s | 55.7 tok/s | Q4_K_M · 19.3 GB | |
| 3 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 54.5 t/s | 54.5 tok/s | Q4_K_M · 13.1 GB | |
| 4 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 49.1 t/s | 49.1 tok/s | Q4_K_M · 16.5 GB | |
| 5 | Qwen3 30B A3B Thinking 2507Alibaba Q4_K_M · 19.4 GB · 49.1 t/s | 49.1 tok/s | Q4_K_M · 19.4 GB |
The Mac Mini M4 Pro is a 48 GB mini PC with more memory than speed: it runs 14B and 35B MoE models well. macOS keeps about 3.5 GB, leaving 44.5 GB for models. Prompt reading is its weak spot: a 4K-token prompt waits 11 s before the first word on Qwen3.6 35B A3B. At its $2,199 MSRP it gives 13.6 tok/s per $1,000, 5th of the 6 machines compared here. M4 Pro generation, released 2024; the newer M5 Pro writes 13% faster with 1.9× the compute.
Mac Mini M4 Pro strengths and weaknesses
Strengths
- Runs 35B MoE models: Qwen3.6 35B A3B at 43.2 tok/s in 22 GB
- Low power draw: 140 W at full load for the whole machine
Weaknesses
- 120B MoE models don't fit: GPT OSS 120B needs 70.8 GB at Q4 vs 44.5 GB usable here
- Slow on large dense models: Llama 3.3 70B Instruct at 4.8 tok/s, Qwen3.8 27B at 11.0 tok/s
- Slow prompt reading: A 32K-token prompt on Qwen3.6 35B A3B waits 1.5 min for the first token
- No CUDA: vLLM, AWQ, GPTQ and FP8 builds don't run; use GGUF or MLX
Memory for models
44.5 GB of 48 GB120B MoE models need 70.8 GB at Q4, more than the 44.5 GB available.Or rent an RTX PRO 6000 Blackwell Max-Q Workstation Edition (96 GB) from $1.14/hr →
Speed per dollar and per watt
On Qwen3.5 9B (9B params) at Q4_K_M. New means a shop price, either a street quote or the maker's own price; used is compared with used. The last two columns are the speed you get per $1,000 of price and per 100 W of power; higher is better. The Mac Mini M4 Pro has no current new price, so it is shown at its MSRP.
| Machine | New price | tok/s | tok/s per $1,000 | tok/s per 100 W |
|---|---|---|---|---|
Mac Studio M3 Ultra96 GB | about $3,999 | 90.1 | 22.5 | 33.4 |
Mac Mini M4 Pro24 GB | about $1,599 | 30.0 | 18.8 | 21.4 |
| about $3,999 | 60.1 | 15.0 | – | |
| about $2,299 | 33.8 | 14.7 | – | |
Mac Mini M4 Pro48 GB | about $2,199 | 30.0 | 13.6 | 21.4 |
Mac Mini M432 GB | about $999 | 13.2 | 13.2 | 20.3 |
| Machine | Used price | tok/s | tok/s per $1,000 | tok/s per 100 W |
|---|---|---|---|---|
RTX 409024 GBBest used value | about $3,000 | 103 | 34.3 | 22.9 |
Mac Studio M3 Ultra96 GB | about $4,525 | 90.1 | 19.9 | 33.4 |
| about $2,109 | 33.8 | 16.0 | – |
Cost per token: the Mac Mini M4 Pro against 24–96 GB machines and GPUs
On Qwen3.5 9B (9B params, Q4_K_M, 4K context). Speed, cost per token and power per token depend on the model.
- Machine, MSRP
- $4.65
- Power
- $0.19
Renting costs as much as buying after ≈2,200 hours (≈1.5 years at 4 h a day).
Each dot is a machine or GPU; hover or tap it for its numbers, click to open it. Machines without a price for the chosen basis are left out. Machines without a published power figure are left out of Per watt and costed without power. How we calculate cost per token
Other memory sizes of the same machine first, then machines and GPUs people weigh against it, each with the one reason it is worth a look. Speeds on Qwen3.5 9B.
Output speed depends on memory bandwidth (273 GB/s). Grey rows need more than the 44.5 GB available.
- 9BQwen3.5 9B · Q4_K_M · 6.6 GB30.0tok/s
- 14BQwen3 14B · Q4_K_M · 11.9 GB20.1tok/s
- 27–31BQwen3.8 27B · Q4_K_M · 17.9 GB11.0tok/s
- 35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB43.2tok/s
- 4.8tok/s
- 120B MoEGPT OSS 120B · not at 4-bitneeds 70.8 GB
Estimated speeds: Q4_K_M, 4K context, one user. How we estimate
Pro generations: writing vs compute
M1 Pro → M5 ProWriting speed follows memory bandwidth and barely moved: 22.0 to 33.8 tok/s since the M1 Pro. Compute, which sets how fast prompts are read, grew 3.1×. Indexed to the M1 Pro; speeds on Qwen3.5 9B, same 32 GB.
All 1215 models on the Mac Mini M4 Pro
Featured onlyBest quant that fits in 44.5 GB, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works
| Model | Quant | VRAM | Speed | Context | Score | Fit |
|---|---|---|---|---|---|---|
Qwen2.5 0.5B InstructChatFunctions | Q4_K_M | 0.8 GB | 308.2 t/s | 33K | S100, Runs great | |
Gemma 3 4B ITVision | Q4_K_M | 5 GB | 67.3 t/s | — | S85, Runs great | |
Qwen3 VL 2B ThinkingVisionFunctions | Q4_K_M | 3.5 GB | 105.6 t/s | 262K | S92, Runs great | |
C4ai Command R7b 12 2024ChatFunctions | Q4_K_M | 7.9 GB | 36.1 t/s | — | A73, Runs well | |
Gpt2 LargeChat | Q4_K_M | 0.5 GB | 353.9 t/s | 1K | S100, Runs great | |
Qwen3.5 2B BaseVision | Q4_K_M | 1.9 GB | 108.0 t/s | 262K | S93, Runs great | |
Bloomz 560MChat | Q4_K_M | 0.4 GB | 516.5 t/s | 2K | S100, Runs great | |
| Q4_K_M | 0.3 GB | 637.0 t/s | 1K | S100, Runs great | ||
| Q4_K_M | 0.1 GB | 3185.0 t/s | 2K | S100, Runs great | ||
OlmOCR 2 7B 1025Vision | Q4_K_M | 6.2 GB | 35.5 t/s | 128K | A73, Runs well | |
Gemma 3 1B ITChat | Q4_K_M | 2.1 GB | 289.5 t/s | 33K | S100, Runs great | |
Yi 1.5 9B ChatChat | Q4_K_M | 6 GB | 32.9 t/s | 4K | A72, Runs well | |
Qwen3.5 35B A3B BaseVision | Q4_K_M | 22.0 GB | 43.2 t/s | 262K | A77, Runs well | |
Mamba 130M HFChat | Q4_K_M | 0.9 GB | 2123.3 t/s | — | S100, Runs great | |
Nemotron Cascade 2 30B A3BChatReasoning | Q4_K_M | 19.3 GB | 45.7 t/s | 262K | A77, Runs well |
Specs
- Chip
- M4 Pro
- CPU cores
- 14
- GPU cores
- 20
- Unified memory
- 48 GB
- For models
- 44.5 GB (macOS keeps 3.5 GB)
- Memory bandwidth
- 273 GB/s
- FP16 (dense)
- 17.2 TFLOPS
- Neural Engine
- 38 TOPS
- Power
- 140 W max
- Released
- 2024-11-08
- MSRP
- $2,199
Mac Mini M4 Pro FAQ
Is the Mac Mini M4 Pro (48 GB) worth it for local LLMs in 2026?
At its $2,199 MSRP it gives 13.6 tok/s per $1,000 on Qwen3.5 9B, 5th of the 6 machines compared here. It runs 14B and 35B MoE models well; dense 27–31B fits but is slow at 11 tok/s, and 70B only just fits in 44.5 GB.
How much memory is available for AI on the Mac Mini M4 Pro (48 GB)?
About 44.5 GB of the 48 GB. macOS keeps roughly 3.5 GB for itself and other apps.
What is the largest model the Mac Mini M4 Pro (48 GB) can run?
Qwen3.6 35B A3B (36.0B) is the largest at 4-bit or better: at Q4_K_M it uses 22.0 GB, 50% of the available memory, and runs at about 43.2 tok/s.
Can the Mac Mini M4 Pro (48 GB) run 120B MoE models?
No. GPT OSS 120B needs 70.8 GB at Q4, more than the 44.5 GB available.
How long does the Mac Mini M4 Pro (48 GB) take to read a long prompt?
On Qwen3.6 35B A3B, about 11 s for a 4K-token prompt and 1.5 min for a 32K-token document before the first word appears.
Is renting a GPU cheaper than buying the Mac Mini M4 Pro?
An A100 (80 GB) rents from $1.00/hr. Renting costs as much as buying after ≈2,200 hours (≈1.5 years at 4 h a day).