Leaderboard/Qwen2.5-Coder-14B-Instruct/RTX 5090 (32GB)/r_opuj21f13-_

Qwen2.5-Coder-14B-Instruct on RTX 5090 (32GB) + AMD Ryzen 7 9850X3D 8-Core Processor (8c) + 30GB

Name: Qwen2.5-Coder-14B-Instruct on RTX 5090 (32GB) + AMD Ryzen 7 9850X3D 8-Core Processor (8c) + 30GB
Creator: llm-speed
License: https://creativecommons.org/licenses/by/4.0/
Keywords: LLM benchmark, tokens per second, Qwen2.5-Coder-14B-Instruct, RTX 5090 (32GB) + AMD Ryzen 7 9850X3D 8-Core Processor (8c) + 30GB, llama.cpp

RTX 5090 (32GB) + AMD Ryzen 7 9850X3D 8-Core Processor (8c) + 30GB

suite suite-v1

cli 0.0.3

signed0tv44ISLy1…

Embed badgesubmitted Jul 01, 2026

https://llm-speed.com/r/r_opuj21f13-_

Workload results

Workload	Backend	Model	decode tok/s	prefill tok/s	TTFT	p50	p95
chat-short	llama.cpp	Qwen2.5-Coder-14B-Instruct	140.0tok/s	—	42.1ms	7.1ms	7.2ms
chat-long	llama.cpp	Qwen2.5-Coder-14B-Instruct	123.0tok/s	—	457ms	8.3ms	8.4ms
concurrent-decode	llama.cpp	Qwen2.5-Coder-14B-Instruct	133.8tok/s	—	—	7.3ms	7.8ms
agent-trace	llama.cpp	Qwen2.5-Coder-14B-Instruct	130.5tok/s	21,985.9tok/s	104ms	7.5ms	8.2ms

Reproduce on your machine

Same workload, same model, signed at your rig. The exact command that produced this run:

$ pipx install llm-speed && llm-speed bench --model qwen2-5-coder-14b-instruct --workload 'chat-short'

Runs in about a minute. Your number lands on the leaderboard signed and linkable. How it's measured.

Embed this run

Drop the badge into a README, blog post, or signature. Each render is a backlink to the signed result.

[![llm-speed: 140 tok/s on RTX 5090 (32GB) (Qwen2.5-Coder-14B-Instruct)](https://llm-speed.com/badge/r_opuj21f13-_.svg)](https://llm-speed.com/r/r_opuj21f13-_)

Related benchmarks

More Qwen2.5-Coder-14B-Instruct benchmarks — every backend and rig that has run this model.
More RTX 5090 (32GB) LLM benchmarks — every model measured on this hardware.

Provenance

Run ID: r_opuj21f13-_
Fingerprint hash: 013ca61a09d17996
Public key: 0tv44ISLy10gz6Oc6FJig3eWJVLwso9oKemSM4nicKM=
Received: 2026-07-01 08:29:56