IBM has released Granite 4.1, a family of open-source language models designed for enterprise use, featuring three sizes (3B, 8B, and 30B parameters). Notably, the 8B dense model demonstrates performance matching or exceeding the previous 32B MoE model across various benchmarks, including ArenaHard and GSM8K. This improvement is attributed to IBM's focus on data quality and a sophisticated multi-phase training process involving 15 trillion tokens and iterative data mixture adjustments. AI
Summary written by gemini-2.5-flash-lite from 3 sources. How we write summaries →
IMPACT IBM's new Granite 4.1 models, particularly the efficient 8B version, offer a compelling alternative for enterprises prioritizing performance and cost predictability.
RANK_REASON Release of an open-source model family with detailed performance benchmarks and training methodology.