MetaLlama 4 Scout on Apple M4 Ultra (192GB)
M4 Ultra (192GB) runs Llama 4 Scout at Q8_0 — 5 tok/s due to limited memory bandwidth. Slow but functional — see all options below.
Model Size
109B
Device VRAM
192 GB
Bandwidth
819 GB/s
Quantization
Q8_0
Performance by Quantization
OwnRig currently has one published compatibility entry for Llama 4 Scout on Apple M4 Ultra (192GB) at Q8_0. This is the best supported pairing we can stand behind today.
| Quantization | Speed | TTFT | Fits in VRAM | Rating | Confidence |
|---|---|---|---|---|---|
| Q8_0 | 5 tok/s | 3040ms | ✓ Yes | Marginal | estimated |
Notes
Q8_0
Estimated from llama-3-3-70b benchmarks. MoE architecture improves throughput vs. dense equivalent.
About Llama 4 Scout
Llama 4 Scout (109B) is a chat, reasoning, multi-purpose model. Large MoE model (~109B total, ~17B active per token; 16 experts, 2 active). Multimodal (text and image) with an enormous advertised context window (10M tokens); practical local runs are usually capped by VRAM and tooling far below that. MoE loads the full weight set for common local inference paths. Not a consumer GPU model: even Q4_K_M needs ~60GB VRAM. Expect H100-class or dual A100 hardware for practical deployment. Do not recommend to RTX 4090 or Mac users.
View all Llama 4 Scout hardware options →About Apple M4 Ultra (192GB)
Apple M4 Ultra (192GB) has 192 GB at 819 GB/s. Available in Mac Studio, Mac Pro.
See all models Apple M4 Ultra (192GB) can run →Estimate method: Estimated from comparable model benchmarks. Reference hardware source: github.com (2026-03-14)
Performance varies by driver version, inference engine, quantization method, context length, and system configuration. Figures shown are estimates based on community benchmarks and may not reflect your exact setup. Product names are trademarks of their respective owners. OwnRig is independent and not affiliated with any hardware or AI model provider.