Needs ffmpeg too

llama-server command: llama-server
-hf unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_M
--fit-ctx 128000
-np 1
-fa on
-b 2048
-ub 2048
-ctk q8_0
-ctv q8_0
--temp 0.6
--top-p 0.95
--top-k 20
--min-p 0.0
--presence-penalty 0.0
--repeat-penalty 1.0
--reasoning-budget -1
--chat-template-kwargs "{"preserve_thinking": true}"

llama-server -hf unsloth/Qwen3.5-9B-GGUF:UD-Q6_K_XL --no-mmproj -c 8192 --reasoning-budget 0.5