Canonical Rankings

Best Macs for this model

Qwen 2.5 72B ranked across the Mac lineup at the best practical quantization, using the best available runtime evidence. Historical baseline selected; model picker is focused on current-market choices.

Model

Quantization

Sort

Runtime

29 ranked MacsUse the strongest current runtime evidence for each row.23 other historical models hiddenBaselinesStatic paths cover only canonical model pages; sort and quantization stay as query state.

Historical baseline selected: Qwen 2.5 72B. Default model choices remain current-market; other historical models stay hidden.

Rank	Mac	Score	Quant	Tok/s	Runtime	Fits	Evidence	Price	Why it ranks here
1	Mac Studio M3 Ultra 256GB	311	8bit	15.0 tok/s	Ollama	Fits	Estimated	$7,499	8bit is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 185.3 GB headroom remains at this quantization.
2	Mac Pro M2 Ultra 192GB	247	8bit	15.0 tok/s	Ollama	Fits	Estimated	$6,999	8bit is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 121.3 GB headroom remains at this quantization.
3	Mac Studio M4 Max 128GB	183	8bit	15.0 tok/s	Ollama	Fits	Estimated	$4,499	8bit is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 57.3 GB headroom remains at this quantization.
4	MacBook Pro M4 Max 128GB 16-inch	183	8bit	15.0 tok/s	Ollama	Fits	Estimated	$5,999	8bit is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 57.3 GB headroom remains at this quantization.
5	MacBook Pro M5 Max 128GB 16-inch	163	8bit	10.0 tok/s	Ollama	Fits	Estimated	$5,399	8bit is the current best practical quantization. 10.0 tok/s is estimated from nearby benchmark coverage. 57.3 GB headroom remains at this quantization.
6	Mac Studio M3 Ultra 96GB	151	8bit	15.0 tok/s	Ollama	Fits	Estimated	$3,999	8bit is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 25.3 GB headroom remains at this quantization.
7	Mac Studio M4 Max 64GB	130	6bit	15.0 tok/s	Ollama	Fits	Estimated	$2,999	6bit is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 10.2 GB headroom remains at this quantization.
8	MacBook Pro M4 Max 64GB 16-inch	130	6bit	15.0 tok/s	Ollama	Fits	Estimated	$4,499	6bit is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 10.2 GB headroom remains at this quantization.
9	Mac Mini M4 Pro 48GB	123	MXFP4	15.0 tok/s	Ollama	Fits	Estimated	$1,599	MXFP4 is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 9.0 GB headroom remains at this quantization.
10	MacBook Pro M4 Pro 48GB 14-inch	123	MXFP4	15.0 tok/s	Ollama	Fits	Estimated	$2,499	MXFP4 is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 9.0 GB headroom remains at this quantization.
11	Mac Studio M4 Max 48GB	123	MXFP4	15.0 tok/s	Ollama	Fits	Estimated	$2,499	MXFP4 is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 9.0 GB headroom remains at this quantization.
12	MacBook Pro M4 Pro 48GB 16-inch	123	MXFP4	15.0 tok/s	Ollama	Fits	Estimated	$2,999	MXFP4 is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 9.0 GB headroom remains at this quantization.
13	MacBook Pro M4 Max 48GB 14-inch	123	MXFP4	15.0 tok/s	Ollama	Fits	Estimated	$3,499	MXFP4 is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 9.0 GB headroom remains at this quantization.
14	MacBook Pro M4 Max 48GB 16-inch	123	MXFP4	15.0 tok/s	Ollama	Fits	Estimated	$3,999	MXFP4 is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 9.0 GB headroom remains at this quantization.
15	Mac Studio M4 Max 36GB	95	Q2_K	15.0 tok/s	Ollama	Fits	Estimated	$1,999	Q2_K is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 10.7 GB headroom remains at this quantization.
16	MacBook Pro M4 Max 36GB 14-inch	95	Q2_K	15.0 tok/s	Ollama	Fits	Estimated	$2,999	Q2_K is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 10.7 GB headroom remains at this quantization.
17	MacBook Pro M4 Max 36GB 16-inch	95	Q2_K	15.0 tok/s	Ollama	Fits	Estimated	$3,499	Q2_K is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 10.7 GB headroom remains at this quantization.
18	Mac Mini M4 32GB	91	Q2_K	15.0 tok/s	Ollama	Fits	Estimated	$799	Q2_K is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 6.7 GB headroom remains at this quantization.
19	MacBook Air M4 32GB 13-inch	91	Q2_K	15.0 tok/s	Ollama	Fits	Estimated	$1,499	Q2_K is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 6.7 GB headroom remains at this quantization.
20	MacBook Air M4 32GB 15-inch	91	Q2_K	15.0 tok/s	Ollama	Fits	Estimated	$1,699	Q2_K is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 6.7 GB headroom remains at this quantization.
21	Mac Mini M4 24GB	88	IQ2_XS	15.0 tok/s	Ollama	Fits	Estimated	$599	IQ2_XS is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 4.1 GB headroom remains at this quantization.
22	MacBook Air M4 24GB 13-inch	88	IQ2_XS	15.0 tok/s	Ollama	Fits	Estimated	$1,299	IQ2_XS is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 4.1 GB headroom remains at this quantization.
23	Mac Mini M4 Pro 24GB	88	IQ2_XS	15.0 tok/s	Ollama	Fits	Estimated	$1,399	IQ2_XS is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 4.1 GB headroom remains at this quantization.
24	MacBook Air M4 24GB 15-inch	88	IQ2_XS	15.0 tok/s	Ollama	Fits	Estimated	$1,499	IQ2_XS is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 4.1 GB headroom remains at this quantization.
25	MacBook Pro M4 Pro 24GB 14-inch	88	IQ2_XS	15.0 tok/s	Ollama	Fits	Estimated	$1,999	IQ2_XS is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 4.1 GB headroom remains at this quantization.
26	MacBook Pro M4 Pro 24GB 16-inch	88	IQ2_XS	15.0 tok/s	Ollama	Fits	Estimated	$2,499	IQ2_XS is the current best practical quantization. 15.0 tok/s is estimated from nearby benchmark coverage. 4.1 GB headroom remains at this quantization.
27	Mac Mini M4 16GB	0	F32	—	Ollama	No	Estimated	$499	Qwen 2.5 72B does not fit on Mac Mini M4 16GB at the current practical quantization.
28	MacBook Air M4 16GB 13-inch	0	F32	—	Ollama	No	Estimated	$1,099	Qwen 2.5 72B does not fit on MacBook Air M4 16GB 13-inch at the current practical quantization.
29	MacBook Air M4 16GB 15-inch	0	F32	—	Ollama	No	Estimated	$1,299	Qwen 2.5 72B does not fit on MacBook Air M4 16GB 15-inch at the current practical quantization.

Qwen 2.5 72B — ranking first, raw rows below

Start with the ranked Mac table above. Use the rest of this page to inspect raw Apple Silicon coverage and model metadata.

Quantizations observed: Q4_K - Medium

2Benchmark rows

2Chip tiers covered

15.0Fastest avg tok/s (M4 Ultra (192 GB))

—Minimum RAM observed

What this page answers best

Fastest published result is 15.0 tok/s on M4 Ultra (192 GB) at Q4_K - Medium. Published runtimes include Ollama. Start with Rankings for the decision, then use the raw rows below to audit the evidence.

Evidence state: 2 linked reference rows and no Silicon Score Lab rows yet.

Published runtimes here: Ollama.

Need the best Mac for this model? Use Buy Need a setup-first answer? Use Run Checking whether it fits? Use Fit Browse Macs by exact hardware Need the full audit trail? Use Bench Comparing against rented GPUs? Use AI Datacenter Index

Catalog record

72.7BTotal params

DenseActive params

131,072Context window

2024-09-19Release date

This is a reference-only model record. It remains useful for historical benchmarks, migration checks, and audit context, but it is excluded from current frontier packs.

What this model is, and what Apple Silicon users are actually seeing

Official model cards tell you what the model is for and which software stacks it targets. Field reality below shows how much Apple Silicon evidence we have so far.

Official brief

Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.

Official source · Raw model card

codingreasoning

Runtime support mentioned

vLLMTransformers

Official specs

Type: Causal Language Models.
Scale: 72.7B.
Context: Full 131,072 tokens and generation 8192 tokens.
Architecture: transformers with RoPE, SwiGLU, RMSNorm, and Attention QKV bias.
Total parameters: 72.7B.
Max input: Full 131,072 tokens and generation 8192 tokens.

Official takeaways

Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.
Please refer to our Documentation for usage if you are not familar with vLLM.
To handle extensive inputs exceeding 32,768 tokens, we utilize YaRN, a technique for enhancing model length extrapolation, ensuring optimal performance on lengthy texts.
Qwen2.5 brings the following improvements upon Qwen2: This repo contains the instruction-tuned 72B Qwen2.5 model, which has the following features: Type: Causal Language Models Training Stage: Pretraining & Post-trainin…

Official model cards describe intent, capabilities, and supported stacks. They do not prove Apple Silicon speed by themselves.

Field reality on Apple Silicon

2Benchmark rows

0Field reports

0Practitioner signals

Sparse BenchmarksEvidence status

What would improve confidence

Expand Cross Chip Benchmark Coverage
Upgrade To First Party Measurement

Current published coverage

Published chip coverage includes M4 Ultra (192 GB), M5 Max (128 GB). Fastest published row is 15.0 tok/s on M4 Ultra (192 GB) at Q4_K - Medium.

M4 Ultra (192 GB)M5 Max (128 GB)

Related Qwen 2.5 models with published pages: Qwen 2.5 14B · Qwen 2.5 7B

Raw benchmark rows for Qwen 2.5 72B

Rows stay below the ranking because this page is answer-first. Use them to inspect exact chips, quantizations, runtimes, and sources.

Chip	Quant	RAM req.	Context	Avg tok/s	Prompt tok/s	Runtime	Source
M4 Ultra (192 GB)	Q4_K - Medium	—	—	15.0 tok/s	—	Ollama	ref
M5 Max (128 GB)	Q4_K - Medium	—	—	10.0 tok/s	—	Ollama	ref

Chips with published results for Qwen 2.5 72B

M4 Ultra (192 GB)M5 Max (128 GB)

Data

benchmarks.json — full dataset · models.json — model summaries · benchmarks.csv — CSV export

See all models →