Can the Apple M4 Pro (48 GB) run gemma-4-26B-A4B-it?
gemma-4-26B-A4B-it·25.8B (4B active)·Apple M4 Pro (48 GB)·48 GB unified memory
Yes. The Apple M4 Pro (48 GB) runs gemma-4-26B-A4B-it fully in unified memory at Q8_0 (about 29 GB of 36 GB), at roughly 35 tokens/s.
- best quant
- Q8_0
- needs
- 29 GB
- usable
- 36 GB
- est. speed
- ~35 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the gemma-4-26B-A4B-it page →
Every quantisation of gemma-4-26B-A4B-it on the Apple M4 Pro (48 GB)
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 11 GB | fits | ~114 tok/s |
| Q3_K_M | 15 GB | fits | ~77 tok/s |
| Q4_K_M | 18 GB | fits | ~62 tok/s |
| Q5_K_M | 20 GB | fits | ~53 tok/s |
| Q6_K | 23 GB | fits | ~46 tok/s |
| Q8_0 pick | 29 GB | fits | ~35 tok/s |
| FP16 / BF16 | 52 GB | too big | — |
Speed is a bandwidth-based estimate for generating text (273 GB/s); prompt processing and long contexts are slower.
Measured speed on the Apple M4 Pro (48 GB)
No community reports for this pair yet.
Report your own speed →Run gemma-4-26B-A4B-it on the Apple M4 Pro (48 GB)
Using Q8_0, the best fit for Apple M4 Pro (48 GB).
ollama run gemma4:26bThe library build picks its own default quantisation.
ollama run hf.co/unsloth/gemma-4-26B-A4B-it-GGUF:Q8_0 llama-server -hf unsloth/gemma-4-26B-A4B-it-GGUF:Q8_0brew install llama.cpp ships with Metal. Downloads the GGUF on first run.
lms get unsloth/gemma-4-26B-A4B-it-GGUF@Q8_0Then load it from the LM Studio app or with `lms load`.
from transformers import pipeline
pipe = pipeline("image-text-to-text", model="google/gemma-4-26B-A4B-it", dtype="auto", device_map="auto")
messages = [
{
"role": "user",
"content": [
{"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
out = pipe(text=messages, max_new_tokens=256)
print(out[0]["generated_text"][-1]["content"])Vision models that fit the Apple M4 Pro (48 GB) comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 27.8B | Q8_0 | ≈31 GBFits in memory | ~5.1 tok/s | |
| 27.8B | Q8_0 | ≈30 GBFits in memory | ~5.1 tok/s | |
| 31.3B | Q6_K | ≈33 GBFits in memory | ~5.8 tok/s | |
| 27.4B | Q8_0 | ≈32 GBFits in memory | ~5.2 tok/s | |
| 29.8B | Q8_0 | ≈30 GBFits in memory | ~4.7 tok/s | |
| 27.8B | Q8_0 | ≈30 GBFits in memory | ~5.1 tok/s |
Other GPUs that run gemma-4-26B-A4B-it
Frequently asked questions
- Can the Apple M4 Pro (48 GB) run gemma-4-26B-A4B-it?
- Yes. The Apple M4 Pro (48 GB) runs gemma-4-26B-A4B-it fully in unified memory at Q8_0 (about 29 GB of 36 GB), at roughly 35 tokens/s.
- How fast does gemma-4-26B-A4B-it run on the Apple M4 Pro (48 GB)?
- An estimate from memory bandwidth gives about 35 tokens/s at Q8_0.
- Which quantisation of gemma-4-26B-A4B-it should I use on the Apple M4 Pro (48 GB)?
- Q8_0: the largest one that fits. Expected quality: lossless compared with the full-precision weights.
Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.