← sources

Flint anyone?

source: reddit (reddit:LocalLLaMA)published 2026-11-18 10:54ingested 2026-11-19 02:39

retained text

I see a lot of lament about Qwen not making 3.8 35B (and I'm sad about it too) but I dont see anyone discuss Flint 40B, a new coding-oriented MoE model that seems to be atleast on par with qwen 3.6 35B, maybe a bit better on agentic tasks and coding. Very fast in my setup.

I'm still testing it myself, but I wonder if its not liked here or people just arent aware it exists? It performs admirably for me so far, my local model of choice for when I want good speeds. QFN is better ofcourse but for me that means low quants or low speeds.

Curious about you guys' experience with the Flint model. 

document fetched 2026-11-19 00:43 · canonical https://reddit.com/r/LocalLLaMA/comments/1xzcbnc/flint_anyone