llama-modchip
π Run 284B DeepSeek-V4-Flash (<59GiB RAM, 44.4 tok/s pp, 9.1-14.7 tok/s) & 30B MoEs (25-111 tok/s) on a pure 64GB CPU at 256K ctx (0 GPUs). SwiGLU-H2, Stockfish-PVS L3 & 13ms Claude Code restore.
llama-modchip launched on 29 October 2026. It ranked #112 on Daily Launch on 17 November 2026 with a Launch Score of 11, from +39 stars on 17 Nov on GitHub. It moved up 8 places from the day before.
What is Llama Modchip?
Submit your terminal scorecard via a GitHub Benchmark Issue or GitHub Discussions to add your CPU to the leaderboard below.
From its README on GitHub
Llama Modchip (llama-modchip) is a repo by synchopate on GitHub. It is written in Python and released under the Apache-2.0 license.
How to get Llama Modchip
curl -sL https://raw.githubusercontent.com/synchopate/llama-modchip/main/install.sh | bashFrom its README on GitHub
More steps and the source are in its repository on GitHub.
In its own words
Run 284B DeepSeek-V4-Flash (<59GiB RAM: 41.1 cold pp, 9.1 N=1 tg, up to 25.3 spec tok/s) & 30B MoEs (22.4 N=1 / 111.6 spec tok/s) on 64GB CPU (0 GPUs).
From its GitHub repository