Measured model
Measured on a Mac Studio (M5 Max, 36GB)Ministral-3-14B-Instruct-2512
Mistral · Ollama name hf.co/mistralai/Ministral-3-14B-Instruct-2512-GGUF:Q4_K_M
On a Mac Studio (M5 Max, 36GB), Ministral-3-14B-Instruct-2512 answers a simple question in about 0.1 s, gets 72% of the Japanese knowledge questions right, 83% of the math problems, and uses 9.9GB of memory.
These numbers are from before the method changed on Oct 2, 2026 (see how it was measured).
Results
| Measure | Result | Among measured models | Note |
|---|---|---|---|
| Reply time | about 0.1 s | 5 of 30 | Until one simple common-sense question is answered (median of 50) |
| Knowledge (high school to university) | 72% | 19 of 30 | 200 questions, ±7 |
| Math (word problems) | 83% | 11 of 30 | 100 questions, ±9, 2 cut off (counted wrong) |
| Memory used | 9.9GB | 14 of 30 | |
| EN→JA translation | 31 pts | 17 of 28 | out of 100 (chrF), about 1 s a sentence |
| JA→EN translation | 56 pts | 13 of 28 | out of 100 (chrF), about 0.7 s a sentence |
"Among measured models" is the place among the 30 models measured on this site (lower is better for time and memory).
Try it on your Mac
- Install the Mac version of Ollama from ollama.com and start it.
- Type this in Terminal. The first time, the model downloads; then you can chat with it.
ollama run hf.co/mistralai/Ministral-3-14B-Instruct-2512-GGUF:Q4_K_M
It used 9.9GB here. To use it comfortably alongside other apps, a Mac with 24GB of memory or more is a good guide (the model then takes about half the memory).
About the model
- Maker
- Mistral
- Size
- 13.5B parameters
- Quantization
- Q4_K_M
- Thinks before answering
- No
- Longest input at once
- 262,144 tokens
- Released (Hugging Face)
- Oct 31, 2025
- Generation speed
- 45.4 tok/s
- Start-up (loading)
- about 1.1 s
- Measured on
- Sep 29, 2026
- Ollama
- 0.34.3
- Hugging Face
- mistralai/Ministral-3-14B-Instruct-2512
- Series
- Mistral
Models of similar memory
- NVIDIA-Nemotron-Nano-12B-v2 (8.1GB of memory, knowledge 79%, reply about 7.1 s)
- gpt-oss:20b (12.8GB of memory, knowledge 81%, reply about 1.9 s)
- GLM-4-9B-0414 (6.7GB of memory, knowledge 66%, reply about 0.1 s)
- DeepSeek-R1-0528-Qwen3-8B (6.4GB of memory, knowledge 73%, reply about 9.8 s)
- Ornith-1.5-9B (6.4GB of memory, knowledge 78%, reply about 2.5 s)
Compare the main models on the text page. The series page lists every model measured.