Machine
RTX 3080 Ti · Ryzen 7 5800XT #a1fe7b5e
29 runs of 22 models since Nov 22, 2026. Last run Nov 22, 2026.
29
runs
22
models
24
configs
5
configs run 2+ times
Hardware
- GPU
- NVIDIA GeForce RTX 3080 Ti 12 GB
- CPU
- AMD Ryzen 7 5800XT 8-Core Processor 8 cores
- RAM
- 31 GB
- OS
- Manjaro Linux
- Tools
- OpenAI-compatible 26 · llama.cpp 3
- Client
- 0.4.99
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
Medians of each model's best config, 22 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.
configrunsqualityeach runspeed
kintsugi-27b
1 run · 1 config
inkling-2-mini-128k
1 run · 1 config
qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3
2 runs · 1 config
★ bestOpenAI-compatible64k ctxthinking default
×2 runs
82.479.9–84.9
steady
91.3 tok/s
90.9–91.8
qwen3.6-35b-256k
1 run · 1 config
qwen3.6-35b-200k
1 run · 1 config
retry-qwen3.6-35b-q4-128k
1 run · 1 config
inkling-2-mini
1 run · 1 config
retry-qwen36-27b-iq3xxs
1 run · 1 config
qwen3.6-35b-128k
1 run · 1 config
deepseek-v4-pro-9b
1 run · 1 config
Compare with another machine
Run history
Nov 22, 2026
18:27qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3×2OpenAI-compatible91.3 tok/s82.417:51spark-2.6-4bOpenAI-compatible144 tok/s58.017:35L0xRE-27b-LowOpenAI-compatible36.1 tok/s68.305:54kintsugi-27bOpenAI-compatible54.4 tok/s84.404:23inkling-2-mini-64kOpenAI-compatible78.3 tok/s68.904:14qwen35-9bOpenAI-compatible73.7 tok/s74.704:02retry-qwen36-27b-iq3xxsOpenAI-compatible6.8 tok/s79.103:21retry-qwen3.6-35b-q4-128kOpenAI-compatible23.6 tok/s80.003:06retry-twil-8b-q8-16kOpenAI-compatible70.6 tok/s71.003:01retry-qwen3-8b-q8-16kOpenAI-compatible81.2 tok/s61.802:46qwen3.6-35b-256kOpenAI-compatible36.0 tok/s81.902:38qwen3.6-35b-200kOpenAI-compatible35.8 tok/s80.702:28qwen3.6-35b-128kOpenAI-compatible35.5 tok/s78.602:16inkling-2-mini-64kOpenAI-compatible79.5 tok/s58.202:08inkling-2-mini-128kOpenAI-compatible72.5 tok/s83.001:55qwen35-9bOpenAI-compatible77.5 tok/s53.101:52qwen3-8b-64kOpenAI-compatible113 tok/s61.201:51inkling-2-mini-uncensoredOpenAI-compatible74.9 tok/s40.201:39deepseek-v4-pro-9bOpenAI-compatible112 tok/s76.301:35twil-8b-fastOpenAI-compatible102 tok/s68.401:30qwen3-8b-fastOpenAI-compatible114 tok/s63.001:29qwen25-coder-7bOpenAI-compatible88.9 tok/s31.201:27qwen25-7bOpenAI-compatible88.3 tok/s51.701:24spark-2.6-4bOpenAI-compatible145 tok/s49.901:03inkling-2-miniOpenAI-compatible79.3 tok/s79.700:47qwen3-8b-fastllama.cpp114 tok/s57.200:45qwen3-8b-fast×2llama.cpp115 tok/s64.4