# Which open-weight LLMs are best at Julia coding?

**URL:** <https://discourse.julialang.org/t/which-open-weight-llms-are-best-at-julia-coding/138944>\
**Category:** General Usage\
**Tags:** question\
**Created:** [August 20, 2026, 7:16am UTC](https://discourse.julialang.org/t/which-open-weight-llms-are-best-at-julia-coding/138944 "2026-08-20T07:16:00Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Niv](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/niv/32/223440_2.png) [@Niv](https://discourse.julialang.org/u/Niv)\
**Post date:** [August 20, 2026, 7:16am UTC](https://discourse.julialang.org/t/which-open-weight-llms-are-best-at-julia-coding/138944/1 "2026-08-20T07:16:00Z")

</div>

Hi all,

We’re setting up a self-hosted LLM for coding, and Julia is our primary language. The usual benchmarks (SWE-bench, Aider) barely cover Julia, so published scores don’t help much.  
I assume that a lot of the fellows here are using such models and I’m not hte first one, so it’d be great to hear what your’e using and why!

Best,  
Niv

---

<div class="post-metadata">

**Author:** ![starsfordummies](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/starsfordummies/32/210387_2.png) [@starsfordummies](https://discourse.julialang.org/u/starsfordummies)\
**Post date:** [August 20, 2026, 8:16am UTC](https://discourse.julialang.org/t/which-open-weight-llms-are-best-at-julia-coding/138944/2 "2026-08-20T08:16:59Z")

</div>

hi, just a couple days ago there was discussion in [this thread](https://discourse.julialang.org/t/which-is-the-best-open-model-for-julia-coding-and-practical-and-still-reasonably-good/138640) which seems pretty much what you’re interested in

---

<div class="post-metadata">

**Author:** ![Niv](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/niv/32/223440_2.png) [@Niv](https://discourse.julialang.org/u/Niv)\
**Post date:** [August 20, 2026, 8:31am UTC](https://discourse.julialang.org/t/which-open-weight-llms-are-best-at-julia-coding/138944/3 "2026-08-20T08:31:30Z")

</div>

Thank you! I’ll have a look now

---

<div class="post-metadata">

**Author:** ![Mason](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/mason/32/2423_2.png) [@Mason](https://discourse.julialang.org/u/Mason)\
**Post date:** [August 20, 2026, 1:08pm UTC](https://discourse.julialang.org/t/which-open-weight-llms-are-best-at-julia-coding/138944/4 "2026-08-20T13:08:16Z")

</div>

My theory is that if an LLM is too dumb to abstract across the relatively minor differences between synatx and semantics in different programming languages, it’s going to be completely hopeless at abstracting across the various **application domains** that these languages are used in.

Julia code is just much much more similar to C++ or Python code than writing a Web server is to writing an aereodynamical simulation.

I wouldn’t pay too much attention to language specific benchmarks, and look at more general programming performance.