open-weight llms

ranked on what the licence file actually says, then on capability and what it takes to run — because in 2026 'open source' in a launch post and the terms in the repo are routinely different documents.

5 listed · 5 researched by us · featured first, then most upvoted

GLM-5.2

we checked thisfree (mit)

the top-ranked open-weight model on the intelligence index, under plain unmodified mit, with a launch post that tells the truth.

#anyone-who-wants-the-str#needs-the-legal-answer-t

DeepSeek V4 Pro

we checked thisfree (mit)

1.6 trillion parameters of mit-licensed coding ability, still shipping as a preview three months after its stable release was due.

#coding#reasoning-workloads-wher

Mistral Large 3

we checked thisfree (apache 2.0)

europe's frontier answer under apache 2.0 — and mistral resisted the temptation to reach for its old research licence.

#teams-who-want-top-tier-

Kimi K2.7 Code

we checked thisfree (modified mit)

a trillion-parameter agentic coder under 'modified mit' — where the modification only bites if you get very big.

#long-running-agentic-eng#for-anyone-comfortably-b

Kimi K3

we checked thisfree (kimi k3 license)

the strongest open-weight model shipped, under the most conditional licence on this page.

#agentic-coding#browsing-work#for-anyone-below-the-rev

this page is deliberately short: a handful of open-weight llms we looked at properly, rather than everything that exists. if you build one that belongs here, a free listing takes two minutes and $50 pins it to the top of this page.

add a tool →