# Thinking Machines releases Inkling 2 Small, a 594B-total MoE model

> Original title: [AINews] Inkling 2 Small: 17A-594B trained in 68 hours on 12T tokens, beating Muse Spark

- Source: Latent Space
- Published: 2026-11-06T10:35:21.000Z
- Bad Influence: https://badinfluence.news/items/ibwev5qlvpxjhtr25dfjk5kc5
- Original: https://latent.space/p/ainews-inkling-2-small-17a-594b-trained-in-68-hours-on-12t-tokens-beating-muse-spark/

## Summary

Thinking Machines released Inkling 2 Small, a 594B-total, 17B-active multimodal mixture-of-experts model, available now on Tinker. Its Apache 2.0 weights are promised for next week after external safety testing, and Artificial Analysis scored it 31 on its Intelligence Index. The source also reports about 150K output tokens per index task and a cost of $1.18 per task.