Machine
RX 9070 9070 Xt 9070 Gre · Ryzen 7 5800X #a6c6f4d2
21 runs of 15 models since Sep 11, 2026. Last run Oct 5, 2026.
21
runs
15
models
18
configs
3
configs run 2+ times
Hardware
- GPU
- AMD Radeon RX 9070/9070 XT/9070 GRE 15 GB
- CPU
- AMD Ryzen 7 5800X 8-Core Processor 8 cores
- RAM
- 46 GB
- OS
- CachyOS
- Tools
- LM Studio 15 · Strata 3 · llama.cpp 3
- Client
- 0.4.59
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, 15 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
qwen3.8-flash-next-iq3_xxs
1 run · 1 config
qwen3.8-27b-gsq-rco
2 runs · 2 configs
tiel-coder-35b-a3b
2 runs · 1 config
★ bestLM Studio256k ctxthinking default
×2 runs
86.085.7–86.2
steady
28.6 tok/s
28.4–28.8
occamy-1.0-Q6_K
1 run · 1 config
qwen3.8-flash-next-iq2_xs
1 run · 1 config
qwen3.8-flash-next-coder-iq1_m
1 run · 1 config
dirk-qwen3.8-27b@iq3_xxs
3 runs · 2 configs
★ bestLM Studio120k ctxthinking default
×2 runs
83.582.4–84.6
steady
18.6 tok/s
18.5–18.7
Swift-Qwen3.8-27B-Uncensored-Dynamic-MTP-UD-IQ3_XXS
1 run · 1 config
tiel-coder-35b-a3b-mtp
2 runs · 1 config
★ bestLM Studio256k ctxthinking default
×2 runs
82.882.0–83.7
steady
26.3 tok/s
25.4–27.2
dirk-qwen3.8-27b
1 run · 1 config
Compare with another machine
Run history
Oct 5, 2026
15:21qwen3.8-flash-next-iq3_xxsStrata35.4 tok/s89.206:50qwen3.8-flash-next-coder-iq1_mStrata43.3 tok/s84.0Sep 18, 2026
18:08Swift-Qwen3.8-27B-Uncensored-Dynamic-MTP-UD-IQ3_XXSllama.cpp32.3 tok/s82.917:38Qwen3.8-27B-GSQ-RCO-IQ3_XXS-mtpllama.cpp32.0 tok/s81.509:53occamy-1.0-Q6_Kllama.cpp30.4 tok/s85.9Sep 15, 2026
14:56swift-qwen3.8-27b@iq2_xxsLM Studio33.1 tok/s66.410:20qwen3.8-27b-gsq-rcoLM Studio31.7 tok/s88.807:53qwen3.6-35b-a3b-mtpLM Studio27.4 tok/s80.1Sep 14, 2026
16:28qwen3.8-27b-gsq-rcoLM Studio16.5 tok/s87.814:01tiel-coder-35b-a3b-mtpLM Studio27.2 tok/s83.713:46dirk-qwen3.8-27b@iq3_xxsLM Studio18.5 tok/s82.413:29tiel-coder-35b-a3b-mtpLM Studio25.4 tok/s82.0