Machine

2× Gtx 1070 · Core i7-9700KF #1e23084c

8 runs of 3 models since Nov 18, 2026. Last run Nov 20, 2026.

ShareRedditX
8
runs
3
models
5
configs
2
configs run 2+ times

Hardware

GPU
2× NVIDIA GeForce GTX 1070 16 GB
CPU
Intel(R) Core(TM) i7-9700KF CPU @ 3.60GHz 8 cores
RAM
11 GB
OS
Debian GNU/Linux 13 (trixie)
Tools
llama.cpp 8
Client
0.4.99
All 2× NVIDIA GeForce GTX 1070 benchmarks

Best config per model

sort

Runs with the same model, quant, tool, context, thinking and KV cache count as one config. Numbers are the median of its runs, and each dot is one run on the 0–100 quality scale. ★ marks the best config with at least 2 rated runs.

Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3
2 runs · 1 config
★ bestllama.cpp64k ctxthinking on
×2 runs
83.480.3–86.5
steady
20.4 tok/s
20.4–20.5
occamy-1.0-abliterated-PXQN3
4 runs · 2 configs
★ bestllama.cpp64k ctxthinking on
×3 runs
82.980.8–84.0
steady1 slow run
19.5 tok/s
12.4–23.9
Huihui-Qwen3.6-35B-A3B-abliterated-ggml-model-PXQN3
2 runs · 2 configs
★ bestllama.cpp16k ctxthinking on
80.6
1 rated run
17.9 tok/s

Compare with another machine

Run history