toktapethe record, not a recording of it

NVIDIA RTX A6000

81 tapesbloomery53 llama-server17ik_llama.cpp7openai3 exllamav31 qwen3.6-35b-a3b10 qwen2.5-7b-instruct1 qwen3-coder-next-512x2.5b1

Narrowed toquant: iq4_xs ×engine: ik_llama.cpp ×Clear all

1 run, newest first

Qwen3-Coder-Next IQ4_XS · 1 stream · long answer 129 tok/sprefill 1821 tok/s

qwen3-coder-next-512x2.5b

ik_llama.cpp7cd62a3eIQ4_XS80B · 4.2B activeP6881 · G5061 · cache 0%ctx 32kfa onrtx-3090rtx-a6000linux 1 stream ! 2 caveats

midagedev

2026-09-20

Published runs · every tape on this page was recorded on this card and carries its own caveats; the page does not rank them. toktape publish <run.toktape> puts one here.

Page views are counted with Cloudflare Web Analytics: no cookies, no fingerprinting. The toktape binary itself sends nothing anywhere.