IBM releases Granite 4.1, open-source AI model family; 8B dense model outperforms prior 32B version across benchmarks after five-phase training on 15 trillion tokens with obsessive data quality filtering.

Granite 4.1: IBM's 8B Model Is Competing With Models Four Times Its Size - Firethering

- Advertisement - IBM just released Granite 4.1, a family of open source language models built specifically for enterprise use. Three sizes, Apache 2.0 licensed and trained on 15 trillion tokens with a level of pipeline obsession that’s worth understanding. But there’s one result in the benchmarks I keep coming back to. The 8B model. Dense architecture, no MoE tricks, no extended reasoning chains. It matches or beats Granite 4.0-H-Small across basically every benchmark they ran. That older model...