Inkling 2 mini: 26.4B MoE with 3.4B active matches Qwen3 27B under Apache 2.0

LiTianleli · x · 2026-11-21

An Artificial Analysis leaderboard screenshot shows Inkling 2 mini — a MoE model with 26.4B total and only 3.4B active parameters — scoring on par with Qwen3.8 27B while running significantly faster. It is released under the permissive Apache 2.0 license.

Related event: Thinking Machines' Open-Source Inkling 2 Mini Matches Frontier Models at a Fraction of Cost(20 posts)→

Original post →

More from Models

Models channel →