Can I run Gemma 4 26B-A4B on a Mac?

Yes: Gemma 4 26B-A4B runs on 43 of the 54 Macs in the LLMCheck index. It needs about 18 GB at a 4K context, 20 GB at 64K. Smallest and cheapest current Mac at 15+ tok/s is the Mac mini M6 24 GB ($1,099; 49 tok/s community-measured), raised GPU limit. Best measured figure: 120.6 tok/s on an M5 Max (community).

RAM at 4K
18
GB
RAM at 64K
20
GB, 16-bit KV
Macs that run it
43
of 54 configs
Best measured
community120.6
tok/s · M5 Max

Gemma 4 26B-A4B on every Mac

One row per Mac configuration, grouped by product line. The build is the one each speed is for; the context column is the longest prompt that fits inside macOS’s default GPU memory limit with a 16-bit KV cache (“raised” where the limit has to be raised first; “full” when the whole window fits). Speeds are decode tok/s from the LLMCheck index, each with its provenance mark. For small mixture-of-experts models (40B total or less), the formula’s typical error is about 15% (the median gap to 88 measured plain-decode runs; it tends to undershoot (median −7%), and an estimate scaled from the model’s own measured run is usually closer; method).

Runs fits macOS’s default GPU limit, 15+ tok/sRuns, slowly fits, under 15 tok/sRaise GPU limit fits once the limit is raised (command below)Doesn't fit needs more memory than the Mac has
MacVerdictBuildSpeedMax contextModel page
MacBook Air
MacBook Air M5 16 GBDoesn't fit————
MacBook Air M5 24 GBRaise GPU limit command ↓Q4_K_Mestimated48 tok/s99K raised—
MacBook Air M5 32 GBRunsQ4_K_Mestimated48 tok/s256K (full)—
MacBook Air M1 8 GB refurbDoesn't fit————
MacBook Air M4 16 GB usedDoesn't fit————
MacBook Air M2 16 GB refurbDoesn't fit————
MacBook Air M3 16 GB refurbDoesn't fit————
MacBook Air M1 16 GB refurbDoesn't fit————
MacBook Air M4 24 GB usedRaise GPU limit command ↓Q4_K_Mestimated40 tok/s99K raised—
MacBook Air M2 24 GB refurbRaise GPU limit command ↓Q4_K_Mestimated34 tok/s99K raised—
MacBook Air M3 24 GB refurbRaise GPU limit command ↓Q4_K_Mestimated34 tok/s99K raised—
MacBook Pro
MacBook Pro M6 16 GBDoesn't fit————
MacBook Pro M5 Pro 24 GBRaise GPU limit command ↓Q4_K_Mestimated78 tok/s99K raised—
MacBook Pro M6 24 GBRaise GPU limit command ↓QATcommunity49 tok/s99K raisedM6 page
MacBook Pro M6 32 GBRunsQATcommunity49 tok/s256K (full)M6 page
MacBook Pro M5 Pro 48 GBRunsQ4_K_Mestimated78 tok/s256K (full)—
MacBook Pro M5 Max 64 GBRunsQ4_K_Msourced107 tok/s256K (full)M5 Max page
MacBook Pro M5 Pro 64 GBRunsQ4_K_Mestimated78 tok/s256K (full)—
MacBook Pro M5 Max 128 GBRunsQ4_K_Msourced107 tok/s256K (full)M5 Max page
MacBook Pro M3 Max 64 GB refurbRunsQ4_K_Mestimated86 tok/s256K (full)—
MacBook Pro M3 Max 128 GB refurbRunsQ4_K_Mestimated86 tok/s256K (full)—
Mac mini
Mac mini M6 16 GBDoesn't fit————
Mac mini M5 Pro 24 GBRaise GPU limit command ↓Q4_K_Mestimated78 tok/s99K raised—
Mac mini M6 24 GBRaise GPU limit command ↓QATcommunity49 tok/s99K raisedM6 page
Mac mini M6 32 GBRunsQATcommunity49 tok/s256K (full)M6 page
Mac mini M5 Pro 48 GBRunsQ4_K_Mestimated78 tok/s256K (full)—
Mac mini M5 Pro 64 GBRunsQ4_K_Mestimated78 tok/s256K (full)—
Mac mini M2 Pro 16 GB refurbDoesn't fit————
Mac mini M4 16 GB usedDoesn't fit————
Mac mini M4 Pro 24 GB usedRaise GPU limit command ↓Q4_K_Mestimated72 tok/s99K raisedM4 Pro page
Mac mini M4 24 GB usedRaise GPU limit command ↓Q4_K_Mestimated40 tok/s99K raised—
Mac mini M2 Pro 32 GB refurbRunsQ4_K_Mestimated59 tok/s162K—
Mac mini M4 Pro 48 GB usedRunsQ4_K_Mestimated72 tok/s256K (full)M4 Pro page
Mac mini M4 Pro 64 GB usedRunsQ4_K_Mestimated72 tok/s256K (full)M4 Pro page
iMac
iMac M6 16 GBDoesn't fit————
iMac M6 24 GBRaise GPU limit command ↓QATcommunity49 tok/s99K raisedM6 page
iMac M6 32 GBRunsQATcommunity49 tok/s256K (full)M6 page
Mac Studio
Mac Studio M5 Max 36 GBRunsQ4_K_Mestimated93 tok/s256K (full)M5 Max page
Mac Studio M5 Max 48 GBRunsQ4_K_Msourced107 tok/s256K (full)M5 Max page
Mac Studio M5 Max 64 GBRunsQ4_K_Msourced107 tok/s256K (full)M5 Max page
Mac Studio M5 Ultra 96 GBRunsQ4_K_Mestimated140 tok/s256K (full)—
Mac Studio M5 Max 128 GBRunsQ4_K_Msourced107 tok/s256K (full)M5 Max page
Mac Studio M5 Ultra 256 GBRunsQ4_K_Mestimated140 tok/s256K (full)—
Mac Studio M5 Ultra 512 GBRunsQ4_K_Mestimated140 tok/s256K (full)—
Mac Studio M2 Max 32 GB refurbRunsQ4_K_Mestimated86 tok/s162K—
Mac Studio M1 Ultra 64 GB refurbRunsQ4_K_Mestimated120 tok/s256K (full)—
Mac Studio M4 Max 64 GB usedRunsQ4_K_Mestimated101 tok/s256K (full)—
Mac Studio M2 Max 64 GB refurbRunsQ4_K_Mestimated86 tok/s256K (full)—
Mac Studio M1 Ultra 128 GB refurbRunsQ4_K_Mestimated120 tok/s256K (full)—
Mac Studio M2 Ultra 128 GB refurbRunsQ4_K_Mestimated120 tok/s256K (full)—
Mac Studio M4 Max 128 GB usedRunsQ4_K_Mestimated101 tok/s256K (full)—
Mac Studio M2 Ultra 192 GB refurbRunsQ4_K_Mestimated120 tok/s256K (full)—
Mac Studio M3 Ultra 256 GB usedRunsQ4_K_Mestimated121 tok/s256K (full)M3 Ultra page
Mac Studio M3 Ultra 512 GB refurbRunsQ4_K_Mestimated121 tok/s256K (full)M3 Ultra page
Provenance key
sourced
Published figure — vendor or press review
estimated
community
Linked individual run — blog, forum or submission

Context memory is computed from its config.json (5 full layers (2 kv heads x 512) + 25 sliding layers (8 x 256, window 1024); attention_k_eq_v: K and V counted separately (a runtime that stores the shared tensor once needs half)): about 1.6 GB of 16-bit KV cache at 64K tokens. A q8 cache (llama.cpp -ctk q8_0 -ctv q8_0, Ollama OLLAMA_KV_CACHE_TYPE=q8_0) halves that part. Runtimes also grow a compute buffer with context, so a fit within about 1 GB of the limit is optimistic.

Raising the GPU memory limit

macOS lets the GPU wire only part of unified memory by default (Metal’s recommended working set; M5 and M6 Macs run macOS 26 or later, whose default is higher), and Gemma 4 26B-A4B needs about 18 GB. On the Macs marked Raise GPU limit it fits in RAM but not inside that default. The default is a performance threshold, not a hard wall: past it, the runtime moves part of it to the CPU or shortens the context (llama.cpp, Ollama), runs it slowly (MLX), or stops with an out-of-memory error. One command raises the limit until the next restart (sudo sysctl iogpu.wired_limit_mb=0 restores the default sooner); it leaves macOS a reserve, and other apps get less memory while the model is loaded.

Gemma 4 26B-A4B: memory by quantization

Weights are 26B parameters at each format’s effective size (the index’s GB-per-billion table), plus about 2 GB for the runtime and a short context; longer contexts add the KV cache computed from the model’s config.json at 16 bits. It is a mixture-of-experts model: all 26B parameters sit in memory although only 3.8B are read per token. The first row is the catalog’s own figure (about 18 GB at a 4K context), the one every verdict on this page uses. “Smallest Mac” is the smallest memory size among the Macs Apple sells new that holds it at a 4K context inside that Mac’s default GPU limit. Estimates, not measurements.

BuildWeightsRAM at 4Kat 32Kat 64KSmallest Mac
As catalogued—18 GB19 GB19 GB32 GB (24 GB with a raised limit)
Q2_K8.8 GB11 GB11 GB12 GB16 GB
Q3_K_M12.7 GB15 GB15 GB16 GB24 GB
Q4_K_M14.8 GB17 GB17 GB18 GB24 GB
Q5_K_M18.5 GB20 GB21 GB22 GB32 GB
Q6_K21.3 GB23 GB24 GB25 GB32 GB
Q8_027.6 GB30 GB30 GB31 GB48 GB (36 GB with a raised limit)
F1652.0 GB54 GB55 GB55 GB96 GB (64 GB with a raised limit)

Measured runs of Gemma 4 26B-A4B

Every published (filled square) and community (hollow circle) run the index holds for Gemma 4 26B-A4B: 1 published, 5 community, each linked to its source. LLMCheck runs no tests of its own. One row is one measurement session; on each chip the table above shows the lower median of the eligible runs (plain decode at 5 bits or fewer), and a binned or “not an anchor” run is shown but never used as a reference.

ChipRuntimeQuantDecodeContextDateSource
M3 Ultra 512 GBllama.cppMXFP4_MOEcommunity87.8 tok/sTG32, depth 02026-10-13llamaperf.com not an anchor
M5 Max 128 GBMLXQATcommunity120.6 tok/s1k2026-10-25omlx.ai
M5 Max —llama.cppQ4_K_Msourced107 tok/s4K-token input2026-10-13huggingface.co
M6 32 GBMLX4-bitcommunity52.2 tok/s~530-token prompt2026-09-24github.com
M6 32 GBMLXQATcommunity49.4 tok/s1k2026-10-30omlx.ai
M6 32 GBLM StudioQATcommunity48 tok/s—2026-10github.com

Recommended setup

The quickest start is Ollama. Pull the model by its tag (checked on ollama.com, 2026-11-14; the file is about 19 GB):

ollama run gemma4:26b-a4b

Prefer a window to a terminal? LM Studio downloads the same model with one click. · model card →

On the Mac mini M6 (32 GB), the smallest current Mac that runs it inside the default GPU limit, the index figure is 49 tok/s (community run) at QAT; keep the context at or under 256K there, or use a q8 KV cache. The gemma4:26b-a4b file fits there.

Step-by-step: the Ollama install guide.

🛒 Which Mac to get for Gemma 4 26B-A4B

Mac mini M6 (24 GB) — smallest current Mac that runs it and cheapest at 15+ tok/s, $1,099 new (Apple list price, checked 2026-11-14), 49 tok/s (community run) once the GPU memory limit is raised.

As an Amazon Associate, LLMCheck earns from qualifying purchases. Affiliate links cost you nothing extra and never influence the verdicts.

FAQ: Gemma 4 26B-A4B on a Mac

Can I run Gemma 4 26B-A4B on a Mac?

Yes. Gemma 4 26B-A4B needs about 18 GB of unified memory at a short (4K-token) context, so it runs on 43 of the 54 Mac configurations in the LLMCheck index; 32 of them hold it inside macOS’s default GPU memory limit and 11 only after raising that limit. The smallest Mac Apple sells new that runs it is the Mac mini M6 (24 GB).

How much RAM does Gemma 4 26B-A4B need?

About 18 GB at a 4K context: the weights plus about 2 GB for the runtime. Context adds KV cache: about 19 GB in total at 32K and 20 GB at 64K with the default 16-bit cache (about 19 GB at 64K with a q8 cache). These are estimates from the model’s config.json, not measurements.

How fast does Gemma 4 26B-A4B run on a Mac?

On an M5 Max, the chip the leaderboard quotes it on, the index figure is 107 tok/s (published figure) at Q4_K_M. The best measured figure is 120.6 tok/s on an M5 Max 128 GB (community run, QAT). Every Mac’s figure is in the table on this page with its provenance: published, community run or estimated.

What is the cheapest Mac that runs Gemma 4 26B-A4B well?

Among Macs Apple sells new, the Mac mini M6 (24 GB), $1,099 new (Apple list price, checked 2026-11-14), is the cheapest that reaches 15 tok/s with Gemma 4 26B-A4B, at 49 tok/s (community run), after raising the GPU memory limit.

Do I need to raise the GPU memory limit to run Gemma 4 26B-A4B?

On 11 of the 54 configurations, yes: iMac M6 24 GB, Mac mini M5 Pro 24 GB, Mac mini M6 24 GB and 8 more. macOS lets the GPU use only part of unified memory by default (about 18 GB on the iMac M6 (24 GB)), and Gemma 4 26B-A4B needs about 18 GB. Running sudo sysctl iogpu.wired_limit_mb=20480 in Terminal raises the limit on a 24 GB Mac until the next restart. From 32 GB up, every Mac Apple sells new holds it inside the default limit.

Which quantization of Gemma 4 26B-A4B should I use?

Q4_K_M (about 15 GB of weights) is the common 4-bit GGUF choice. Q8_0 (about 28 GB) is closer to full quality and needs a 48 GB Mac at the default GPU limit. Q3_K_M and Q2_K shrink it to about 13 and 9 GB at a quality cost. The index’s reference figure (107 tok/s on an M5 Max) is for its Q4_K_M build.

How much context can Gemma 4 26B-A4B use on a Mac?

Its own window is 256K tokens. At macOS’s default GPU limit, the Mac mini M6 (32 GB) holds about 256K (the full window).

Cite this page

LLMCheck. (2026). Can I run Gemma 4 26B-A4B on a Mac? llmcheck.net. https://llmcheck.net/can-i-run/gemma-4-26b-a4b/ (updated ). CC BY 4.0.

How to cite LLMCheck →Data: CC BY 4.0Markdown version

Related

Figures from the LLMCheck index, data version 2026-11-18. Verdicts and context limits are estimates from each model’s catalog memory figure and config.json.