Amethyst-13B: benchmark results

Provider: Undi95. Access: Open.

Aggregate Rating 1364 ± 36, rank #7321 of 10916 rated models, from 6 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard v1 - ARC Challenge62.63Normalized accuracy (%) (25-shot)59
Open LLM Leaderboard v1 - HellaSwag83.17Normalized accuracy (%) (10-shot)56.5
Open LLM Leaderboard v1 - TruthfulQA MC252.43MC2 (%) (0-shot)55.8
Open LLM Leaderboard v1 - WinoGrande74.74Accuracy (%) (5-shot)38.5
Open LLM Leaderboard v1 - MMLU55.91Accuracy (%) (5-shot)37.7
Open LLM Leaderboard v1 - GSM8K10.84Accuracy (%) (5-shot)31.9

Interactive version: theaggregate.ai/model?slug=amethyst-13b · How It Works · Data refreshed daily, snapshot 2026-11-08.