Llama 3 8B Slerp Med 262k: benchmark results

Provider: Meta. Access: Open.

Unified ELO 1445 ± 16, rank #1773 of 2928 rated models, from 17 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open Medical LLM - MedMCQA56.37Accuracy (%)66.5
Open Medical LLM - MedQA (USMLE)59.7Accuracy (%)65.6
Open Medical LLM - MMLU Anatomy62.96Accuracy (%)61.6
Open Medical LLM - PubMedQA74Accuracy (%)51.1
Open Medical LLM65.56Average Accuracy (%)50
Open Medical-LLM Leaderboard65.56Medical average (%)48.9
Open LLM Leaderboard v1 - GSM8K27.14Accuracy (%) (5-shot)45.1
Open Medical LLM - MMLU Medical Genetics73Accuracy (%)44.9
Open Medical LLM - MMLU Professional Medicine68.01Accuracy (%)43.2
Open LLM Leaderboard v1 - MMLU56Accuracy (%) (5-shot)38
Open Medical LLM - MMLU College Biology70.83Accuracy (%)37.2
Open Medical LLM - MMLU Clinical Knowledge66.79Accuracy (%)34.1

Interactive version: theaggregate.ai/model?slug=llama-3-8b-slerp-med-262k · How It Works · Data refreshed daily, snapshot 2026-09-23.