Last active
May 23, 2026 14:11
-
-
Save selfup/241448ab853aa284f71d9350194ba002 to your computer and use it in GitHub Desktop.
Ministral 3 (3B, 8B) Instruct - Ubuntu 3800x 3060ti llama.cpp benchmark
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| set -eou pipefail | |
| llama-bench \ | |
| -m ~/.models/Ministral-3-3B-Instruct-2512-Q4_K_M.gguf \ | |
| -m ~/.models/Ministral-3-8B-Instruct-2512-Q4_K_M.gguf \ | |
| -p 512 -n 128 -fa 1 -ngl 99 \ | |
| -d 0,4096,8192,16384 \ | |
| -o md > local-ministral-q4km-${MACHINE}.md 2>/dev/null |
Author
Author
| model | size | params | backend | ngl | fa | test | t/s |
|---|---|---|---|---|---|---|---|
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | pp512 | 5380.04 ± 111.46 |
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | tg128 | 146.67 ± 0.21 |
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | pp512 @ d4096 | 4342.83 ± 49.64 |
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | tg128 @ d4096 | 126.16 ± 0.27 |
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | pp512 @ d8192 | 3555.07 ± 150.68 |
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | tg128 @ d8192 | 112.13 ± 0.13 |
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | pp512 @ d16384 | 2720.28 ± 18.16 |
| mistral3 3B Q4_K - Medium | 1.99 GiB | 3.43 B | CUDA | 99 | 1 | tg128 @ d16384 | 91.19 ± 0.10 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | pp512 | 2516.59 ± 27.66 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | tg128 | 72.50 ± 0.05 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | pp512 @ d4096 | 2203.07 ± 11.29 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | tg128 @ d4096 | 65.56 ± 0.06 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | pp512 @ d8192 | 1943.73 ± 12.84 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | tg128 @ d8192 | 60.38 ± 0.02 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | pp512 @ d16384 | 1573.30 ± 5.67 |
| mistral3 8B Q4_K - Medium | 4.83 GiB | 8.49 B | CUDA | 99 | 1 | tg128 @ d16384 | 51.96 ± 0.03 |
build: b0df4c0cf (9297)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
build: 1e5ad35d5 (9093)