llama-3.2-Korean-Bllossom-AICA-5B — benchmark results

Provider: Meta. Released 2024-12-12. Access: Open.

Unified ELO 1289 ± 46, rank #1555 of 1776 rated models, from 17 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard - IFEval51.72Score61.6
Open LLM Leaderboard - MATH Level 512.39Score55.5
Open LLM Leaderboard - GPQA6.49Score54.5
KOFFVQA - Graph and Chart Understanding50.67Score (%)36.4
KOFFVQA - Korean OCR35Score (%)32.1
KOFFVQA - Document Understanding36.67Score (%)29.6
Open LLM Leaderboard - MMLU-Pro19Score29
Open LLM Leaderboard - MuSR5.82Score28.2
Open LLM Leaderboard - BBH18.65Score27.9
KOFFVQA - Table Understanding23Score (%)25.9
KOFFVQA - Relationship43.33Score (%)22.2
KOFFVQA - Recognition35Score (%)19.8

Interactive version: theaggregate.ai/model?slug=llama-3-2-korean-bllossom-aica-5b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.