← All reports & posts

"Local AI is dead" - 22 local-first repos say otherwise

2026-10-06

The claim: "Local AI is dead" it shows that people are still discovering and paying attention to local-first projects.

We have covered 22 Local-first repositories since 2026-03-10. 16 of them have a star count recorded from the day we covered them. Between those days and 2026-09-13 they gained +39,417 stars (+9.4%).

A star count measures attention, not health, adoption or revenue. It can answer whether people are still arriving; it cannot answer whether the software is good or whether anyone ships it.

So the idea that “Local AI is dead” is too strong. But stars only show interest on GitHub—they don’t tell us how many people actually use these projects, how good they are, or whether they are making money.

What the numbers show

The 16 repositories with a star baseline held 419,336 stars between them on the day each was covered. Today they hold 458,753, a change of +39,417 (+9.4%).

15 gained stars, 1 finished level.

6 of the 22 repositories have no star count recorded from the day they were covered, so no change can be stated for them. They are listed below without a delta. We do not estimate figures we did not capture.

Repository Covered Stars then Stars now Change Language
AprilNEA/OpenLogi 2026-08-08 7,901 20,639 +12,738 (+161.2%) Rust
onyx-dot-app/onyx 2026-06-22 24,802 32,026 +7,224 (+29.1%) Python
usememos/memos 2026-05-12 57,164 62,913 +5,749 (+10.1%) n/a
caddyserver/caddy 2026-06-21 70,312 75,652 +5,340 (+7.6%) n/a
appwrite/appwrite 2026-05-12 54,917 57,349 +2,432 (+4.4%) n/a
appsmithorg/appsmith 2026-07-07 39,192 40,860 +1,668 (+4.3%) n/a
nexu-io/html-anything 2026-07-10 7,517 8,745 +1,228 (+16.3%) HTML
StevenBlack/hosts 2026-07-10 30,132 31,040 +908 (+3.0%) Python
KunAgent/Kun 2026-08-12 5,511 6,308 +797 (+14.5%) TypeScript
Hmbown/CodeWhale 2026-08-03 40,245 40,948 +703 (+1.7%) Rust
QuivrHQ/quivr 2026-07-25 39,232 39,507 +275 (+0.7%) Python
dokku/dokku 2026-06-13 31,934 32,128 +194 (+0.6%) Shell
Universal-Debloater-Alliance/universal-android-debloater-next-generation 2026-09-11 8,999 9,115 +116 (+1.3%) Rust
mokshablr/gander 2026-09-09 885 920 +35 (+4.0%) JavaScript
FrigadeHQ/yap 2026-09-08 381 391 +10 (+2.6%) Swift
VladUZH/qwen-scribe 2026-09-10 212 212 +0 (+0.0%) Python
genspark-ai/genoffice 2026-08-31 not recorded 6,494 no baseline TypeScript
go-gitea/gitea 2026-06-26 not recorded 57,953 no baseline Go
khoj-ai/khoj 2026-03-19 not recorded 37,274 no baseline Python
n8n-io/n8n 2026-03-10 not recorded 204,019 no baseline TypeScript
open-webui/open-webui 2026-03-14 not recorded 151,655 no baseline Python
unslothai/unsloth 2026-04-19 not recorded 76,022 no baseline Python

Pulling away

They solve a clear problem and make a useful idea easy for people to understand and try. They also fit into things developers already care about, which helps them spread through the community.

Barely moved

6 of the 12 repositories we covered more than 30 days ago gained less than 5% since. Anything covered more recently is left out of this comparison, having had little time to move.

How these numbers were collected

"Stars then" is the count recorded by our pipeline on the day the repository was reviewed. "Stars now" is the count from the GitHub API on 2026-09-13. Both are stored per repository in our star history dataset. We did not run or test any of these projects; every figure is a count, not an assessment.

Each repository links to its own page on https://githubtrends.app, which carries the video review and the full coverage record.

Every repository we have covered under this topic is listed on the Local-first page.