Skip to content
llm-speed
← raw run
wrapped·Jul 01, 2026
/r/r_-txe_hiq44n
your run

Qwen3-32B

on RTX 5090
66.6
tok/sdecode
RTX 5090 (32GB) + AMD Ryzen 7 9850X3D 8-Core Processor (8c) + 30GB
rank in tier
157/165RTX 5090 runstop 95%
best workload
concurrent-decode
where the rig flew
slowest workload
chat-long62.3 tok/s
where the rig struggled
backend
llama.cppllama.cpp is the universal backend — broad model support, modest speed ceiling.
faster than
  • starcoder2-15b-4bit on M3 Ultra64.2 tok/s
  • Codestral-22B-v0.1-4bit on M3 Ultra47.5 tok/s
  • Qwen2.5-32B-Instruct-4bit on M3 Ultra34.6 tok/s

This is a shareable view of one signed llm-speed suite-suite-v1 run. Numbers are not edited. The raw record has every workload, the public key, and the fingerprint hash.