Thinking MachinesLaunchAI ModelsNovember 18, 2026

Thinking Machines releases Inkling 2 Mini, a 26B local model

Read original source →huggingface.co

Inkling 2 Mini is a multimodal looped-transformer MoE with ~26B total and 3B active parameters, 1M-token context, and Apache 2.0 weights. Its 12 middle layers run three times, giving 48 logical layer visits from 24 physical ones.

How this story unfolded

2 days · 6 reports · 41 community posts · 47 of 52 shown

  1. Nov 20
  2. Nov 21
  3. Nov 22

More stories today

Open the live feed