Thinking Machines releases Inkling 2 Small (594B/17B MoE) on Tinker; full Apache 2.0 weights planned next week
Thinking Machines Lab's official account announced Inkling 2 Small. The same-day news page says it is a sparse MoE with 594B total and 17B active parameters, with image inputs and variable effort, and an AA Intelligence Index score of 31. It is available for fine-tuning and inference on Tinker now with closed weights; full weights are planned next week under Apache 2.0 after external safety testing. Pre-training ran about 68 hours on GB300 NVL72 racks in native NVFP4.
Why it matters
Official lab account post with concrete size, Tinker availability, and a dated open-weight plan, backed by the same-day news page.
Sources2
Thinking Machines LabMediaIntroducing Inkling 2 SmallView original ·
- @thinkymachinesXIntroducing Inkling 2 Small. Scoring 31 on the AA Intelligence Index, it's a big jump over its predecessor. Inkling 2 Small is available for fine-tuning and inference on Tinker today, with full weight release planned soon. https://thinkingmachines.ai/news/inkling-2-small/View original ·
Written by dalao.ai: the title, summary and why it matters are our own words. The original reports are the record; their publishers keep authorship.