granite-3.2-2B-instruct — benchmark results

IBM's Apache-2.0 2B Granite 3.2 instruct model (February 2025) with an experimental chain-of-thought reasoning mode that can be toggled on or off. Provider: IBM. Released 2025-02-17. Access: Open.

Unified ELO 1405 ± 37, rank #1202 of 1776 rated models, from 7 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard - IFEval61.52Score73.6
Open LLM Leaderboard - MATH Level 514.43Score60.8
Open LLM Leaderboard - GPQA5.37Score44.9
Open LLM Leaderboard - BBH21.67Score32.3
Open LLM Leaderboard - MMLU-Pro19.82Score30.6
Open LLM Leaderboard - MuSR4.7Score23.2
Wolfram LLM Benchmarking Project15.5Correct Functionality (%)11.9

Interactive version: theaggregate.ai/model?slug=granite-3-2-2b-instruct · How the rankings work · Data refreshed daily, snapshot 2026-07-22.