Machine

RTX 3070 Ti · Ryzen 5 5600X #b36c3f2c

9 runs of 6 models since Nov 10, 2026. Last run Nov 11, 2026.

ShareRedditX
9
runs
6
models
7
configs
1
configs run 2+ times

Hardware

GPU
NVIDIA GeForce RTX 3070 Ti 8 GB
CPU
AMD Ryzen 5 5600X 6-Core Processor 6 cores
RAM
47 GB
OS
Windows 11 Pro 25H2
Tools
Polystrata 5 · polystrata 4
Client
0.4.99
All NVIDIA GeForce RTX 3070 Ti benchmarks

Models on this machine

One dot per model, at its best config. Up is better answers, right is faster. Time to finish counts every token the model wrote, thinking included, so a model that talks a lot ends up further left.

Best trade-off: quality 60+, and nothing else here is both faster and betterOther models · dot size = number of runs
below 20 tok/s: slower than you read4050607080901002050tok/s (log scale) →quality →outliers sit at the edges

Medians of each model's best config, 6 models.

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.

ornith-1.5-35b-a3b-apex-i-quality
2 runs · 2 configs
★ bestPolystrata64k ctxthinking default
88.0
1 rated run
42.2 tok/s
swift-1.5-iq2_xs
1 run · 1 config
★ bestpolystrata64k ctxthinking default
86.6
1 rated run
43.4 tok/s
cyber-tiel-coder-35b-a3b-ud-q6_k_xl
1 run · 1 config
★ bestQ6_KPolystrata64k ctxthinking default
85.9
1 rated run
34.0 tok/s
swift-1.5-iq3_xxs
3 runs · 1 config
★ bestpolystrata64k ctxthinking default
×3 runs
84.681.9–85.4
steady
24.8 tok/s
24.0–26.5
qwen3.6-35b-a3b-ud-q4_k_m
1 run · 1 config
★ bestQ4_K_MPolystrata64k ctxthinking on
78.2
1 rated run
78.7 tok/s
tiel-coder-35b-a3b-ud-q4_k_xl
1 run · 1 config
★ bestQ4_KPolystrata64k ctxthinking on
68.8
1 rated run
54.0 tok/s

Compare with another machine

Run history