if you like what I build (taste skill, the demos, all the oss stuff), you can support me directly :)
it would honestly mean a lot. it goes straight into better hardware (this windows laptop is dying on every 3d demo 😭) and the ai subs I need to keep building!
github.com/sponsors/Leonxlnx
also if any company wants to sponsor taste skill or tastecode, hmu :)
here is the repo :)
github.com/Leonxlnx/afterglo…
here is the github repo!
github.com/Leonxlnx/procedur…
I wanted to clarify that this is NOT a paid post (and neither is this one).
And I was wrong: after some testing, it is definitely not better than Opus 5.5, and the rate limits are not good on subscriptions.
It's still a solid model. It doesn't overthink as much as Kimi K3 and brings many improvements, but I don't see myself using it over Opus 5.5.
Holy shit. Kimi K3.1 Pro (max) is like 3x cheaper than Opus 5.5 and even slightly better than Opus 5.5 on Terminal Bench 4.0???
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model.
🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal.
🔹 Full-project coding: games in Unreal, Unity & Godot.
🔹 Excels at parametric 3D modeling in Blender & CAD.
🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%.
🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks.
Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
Open weights by Nov 4, 2026.
🔗 API: platform.kimi.ai
🔗 Tech blog: kimi.ai/blog/kimi-k3-1-pro








