Minds taps MiniMax M3 open-source models to cut Animoca compute costs ~20x

MiniMax_AI · x · 2026-09-15

Minds CEO Yat Siu told Tech in Asia the platform pools 50+ AI models, leaning on open-source LLMs like MiniMax's M3 for 90-95% of frontier capability at a fraction of the cost. This has cut Animoca Brands' compute cost roughly 20x, bringing per-user cost down to cents.

The same economics power Minds' Bazaar, where 5,000+ Skills have been built without heavy technical setup: the first builder covers token costs, subsequent users pay a fraction, and creators earn a revenue cut.

Original post →

More from Venture

Venture channel →