Can the Apple M4 Pro (48 GB) run RWKV7-G1k-13.3B-20260930?
RWKV7-G1k-13.3B-20260930·13.3B·Apple M4 Pro (48 GB)·48 GB unified memory
Yes. The Apple M4 Pro (48 GB) runs RWKV7-G1k-13.3B-20260930 fully in unified memory at FP16 / BF16 (about 34 GB of 36 GB), at roughly 5.7 tokens/s.
- best quant
- FP16 / BF16
- needs
- 34 GB
- usable
- 36 GB
- est. speed
- ~5.7 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the RWKV7-G1k-13.3B-20260930 page →
Every quantisation of RWKV7-G1k-13.3B-20260930 on the Apple M4 Pro (48 GB)
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 12 GB | fits | ~34 tok/s |
| Q3_K_M | 14 GB | fits | ~23 tok/s |
| Q4_K_M | 16 GB | fits | ~19 tok/s |
| Q5_K_M | 17 GB | fits | ~16 tok/s |
| Q6_K | 19 GB | fits | ~14 tok/s |
| Q8_0 | 22 GB | fits | ~11 tok/s |
| FP16 / BF16 pick | 34 GB | fits | ~5.7 tok/s |
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 RWKV7-G1k-13.3B-20260930 on the Apple M4 Pro (48 GB)
from transformers import pipeline
pipe = pipeline("text-generation", model="RWKV/RWKV7-G1k-13.3B-20260930", dtype="auto", device_map="auto")
messages = [{"role": "user", "content": "Explain self-hosting in one paragraph."}]
out = pipe(messages, max_new_tokens=512)
print(out[0]["generated_text"][-1]["content"])Chat models that fit the Apple M4 Pro (48 GB) comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 14.8B | Q8_0 | ≈17 GBFits in memory | ~9.6 tok/s | |
| 14.7B | Q8_0 | ≈17 GBFits in memory | ~9.6 tok/s | |
| 20.9BA4.2B | Q8_0 | ≈13 GBFits in memory | ~34 tok/s | |
| 8.2B | Q8_0 | ≈10 GBFits in memory | ~17 tok/s | |
| 8B | Q8_0 | ≈9.8 GBFits in memory | ~18 tok/s | |
| 9.4B | Q8_0 | ≈11 GBFits in memory | ~15 tok/s |
Other GPUs that run RWKV7-G1k-13.3B-20260930
Frequently asked questions
- Can the Apple M4 Pro (48 GB) run RWKV7-G1k-13.3B-20260930?
- Yes. The Apple M4 Pro (48 GB) runs RWKV7-G1k-13.3B-20260930 fully in unified memory at FP16 / BF16 (about 34 GB of 36 GB), at roughly 5.7 tokens/s.
- How fast does RWKV7-G1k-13.3B-20260930 run on the Apple M4 Pro (48 GB)?
- An estimate from memory bandwidth gives about 5.7 tokens/s at FP16 / BF16.
- Which quantisation of RWKV7-G1k-13.3B-20260930 should I use on the Apple M4 Pro (48 GB)?
- FP16 / BF16: 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.