Back to search results
Ministral-3-8B-Instruct-2512 · IQ3_S
Apple M2 UltraOllama v0.43L2 Cross-framework verified
This page aggregates 1 real-world runs of Ministral-3-8B-Instruct-2512 (IQ3_S) on Apple M2 Ultra with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.
108.26 tok/s
Decode
Decode speed
586.93 tok/s
Prefill
Prefill speed
0.29 s
TTFT
Time to first token
13.19 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
11 days ago
Last verified
Performance
- llama.cpp · Q3_K_M Decode 119.5 · Prefill 580.12 ·
- Ollama · IQ3_S (current)Decode 108.26 · Prefill 586.93 ·
Core figures
- Decode (avg)
- 108.26 tok/s
- Prefill (avg)
- 586.93 tok/s
- TTFT (avg)
- 0.29 s
- VRAM (avg)
- 13.19 GB
- MTP acceptance rate
- —
- TTFB
- 0.87 GB
- Power draw
- 103.1 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Ministral-3-8B-Instruct-2512
- Quantization
- IQ3_S
- Framework
- Ollama
- Version
- v0.43
- Context length
- 16384 tokens
- Batch size
- 256
- GPU layers
- 33
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- Apple M2 Ultra
- Measured VRAM (avg)
- 13.19 GB
- OS
- macOS 26.1
- Power draw
- 103.1 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.43 · macOS 26.1 · 16384 ctx
108.26 tok/s
Decode
586.93 tok/s
Prefill
0.29 s
TTFT
13.19 GB
VRAM
—
MTP
103.1 W
Power draw
L2 batch seed 2026-10-24: Ollama on hw_id=103 running model_id=74 at IQ3_S