Machine

RTX 2060 · Ryzen 7 4800H #cbcc5233

10 runs of 7 models since Sep 22, 2026. Last run Nov 8, 2026.

ShareRedditX
10
runs
7
models
10
configs
0
configs run 2+ times

Hardware

GPU
NVIDIA GeForce RTX 2060 6 GB
CPU
AMD Ryzen 7 4800H with Radeon Graphics 8 cores
RAM
30 GB
OS
Arch Linux
Tools
llama.cpp 10
Client
0.4.99
All NVIDIA GeForce RTX 2060 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 →

Medians of each model's best config, 7 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.

IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF:1
3 runs · 3 configs
★ bestllama.cpp64k ctxthinking on
87.5
1 rated run
25.9 tok/s
peculiar-ragdoll/Cyber-Tiel-Coder-35B-A3B-GGUF
1 run · 1 config
★ bestQ4_K_XLllama.cpp128k ctxthinking default
83.0
1 rated run
25.9 tok/s
IsValorum/Occamy-1.0-APEX-I-MiniPlus-V2.1-Abliterated-GGUF:ABLITERATED
1 run · 1 config
★ bestllama.cpp64k ctxthinking default
80.0
1 rated run
21.7 tok/s
IsValorum/Qwen3.8-35B-A3B-Distill-MTP-APEX-I-MiniPlus-V2.1-Abliterated-GGUF:ABLITERATED
1 run · 1 config
★ bestllama.cpp256k ctxthinking default
79.4
1 rated run
25.7 tok/s
LiquidAI/LFM2.5-24B-A2B-GGUF
2 runs · 2 configs
★ bestQ4_K_Mllama.cpp128k ctxthinking default
77.5
1 rated run
26.8 tok/s
inclusionAI/Ling-3.0-tiny-GGUF
1 run · 1 config
★ bestQ4_K_Mllama.cpp64k ctxthinking default
58.1
1 rated run
55.7 tok/s
ItachiSan/LiquidCode-2.5-24B-A2B-GGUF
1 run · 1 config
★ bestQ4_K_Mllama.cpp128k ctxthinking default
52.4
1 rated run
30.0 tok/s

Compare with another machine

Run history