More
Submit build

Builds / Apple M3 Ultra (unified, Mac Studio)

Mac Studio M3 Ultra 96 GB: Ripple-65B Preview Q4_K_M on llama.cpp at 15.0 tok/s (13.5 after a 32K prompt, 9.45 at 128K)

Last updated:

Source: leanzero.net ↗📏 How measured: reported by the original source; RigBase re-checked the numbers at that source
GPU(s)
Apple M3 Ultra (unified, Mac Studio)
VRAM / unified memory (GB)
96
CPU
Apple M3 Ultra
Motherboard
Mac Studio
Amazon · eBay (used) · Newegg
RAM
96 GB unified
PSU
not stated
Other
Self-made Q4_K_M (llama.cpp b12230 from the official weights, 38.8 GB); mlx-lm 0.32.2 cannot load the architecture yet
Reported total cost
not stated
Power
not stated

Measured speed (generation tokens/sec)

ModelQuantEnginetok/sPrompt processing (tok/s)Details
Ripple 65B PreviewQ4_K_M (self-quantized, 38.8 GB)llama.cpp (Metal)15.0126 @ 512-token prompt
est. TTFT @8k ≈ 65.1 s (calculated from measured prefill)
context: llama-bench, empty context📏 llama-bench tg128; 13.7 with 32K in context, 9.3 at 128K (that row ran with nothing else on the Mac); through llama-server 13.5 after a 32K prompt (43.1 GB peak) and 9.45 after 128K (49.8 GB)

Prompt processing = prefill speed (how fast the prompt is read) as reported by the source, with the prompt length it was measured at; blank = the source does not report it. est. TTFT @8k is calculated from the measured prefill speed (8,192 ÷ prefill tok/s), shown only when prefill was measured on a prompt of 512+ tokens. It is not a measurement and ignores load time and caching. measured TTFT is the source's own time-to-first-token. Why prefill matters for coding assistants.

Owner-run speed test by leanzero (Mihai Perdum, 11 Nov 2026). Caveat from the author: most rows ran while their scorer re-graded a run on the same Mac, so these may slightly understate an idle machine. The same article scores Ripple's agent coding runs (via OpenRouter) low on their Gauntlet test.

Source: https://leanzero.net/blog/ripple-65b-preview-benchmark-lfm3-1-7b

Blank fields = not stated in the source. We never estimate build data; GPU street prices above are separately sourced and dated.

Embed this result

Copy into a README or blog. Links back to this build. See badge docs (CC BY 4.0 attribution). The SVG does not invent speeds — numbers stay on the build page.

Markdown

[![Measured on RigBase](https://getrigbase.com/badge/mac-studio-m3-ultra-96gb-ripple-65b-preview-q4km-llamacpp-leanzero.svg)](https://getrigbase.com/builds/mac-studio-m3-ultra-96gb-ripple-65b-preview-q4km-llamacpp-leanzero/)

HTML

<a href="https://getrigbase.com/builds/mac-studio-m3-ultra-96gb-ripple-65b-preview-q4km-llamacpp-leanzero/" rel="noopener"><img src="https://getrigbase.com/badge/mac-studio-m3-ultra-96gb-ripple-65b-preview-q4km-llamacpp-leanzero.svg" alt="Measured on RigBase" height="20"/></a>

Preview: Measured on RigBase · How to use

Rent before you buy: try this model on a cloud GPU for a few dollars first: Runpod (referral link; sign up with Google to get the new-user bonus)

Quick answers: Can Apple M3 Ultra run Ripple 65B Preview?