Rare find

ChatLLM. Run AI chat models on your computer without the cloud.

github.com/foldl/chatllm.cpp

Problem

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Audience

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Market

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Pricing

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Vaya's read on this project

Problem, audience, market, and the verdict — sign in to see it.

Updates

July 2026

support G9v3 Jul 24
fix neochat (vqa) Jul 21
add doc about lens Jul 16
support OvisOCR2, and clean up. Jul 15
new feature: lens Jul 15
try fix ci Jul 7
fix ci Jul 7
support neochat (vqa) Jul 6

June 2026

support yarn on qwen3.5 Jun 29
upload some models Jun 28
support paddleocr-vl Jun 27
fix some models after fa is added. Jun 25
make `load_tensors_only` faster Jun 22
support mellum-2 Jun 22
convert.py: add `-tt` Jun 15
support HT-MT2 moe Jun 14
initial support of locateanything Jun 12
fix qwen: uses too much mem for video. Jun 10
auto detect multimedia file types Jun 5
support gemma-4-12b Jun 5
`--show`: collapse experts tensors properly Jun 4

May 2026

collapse tensor names May 28
support minicpm5 May 28
some tools for debugging/inspecting. May 27
Release —v2026.05 May 27
upload hy-mt2 dense models. May 27
qwen models: support video input May 26
upload models May 23
support "any" types cross model boundaries. May 22
q4_k reworked: more accurate; less memory. May 21