LongChat-7B-v1.5-32k — benchmark results

Provider: Other. Released 2023-08-01. Access: Open.

Unified ELO 1237 ± 10, rank #1647 of 1776 rated models, from 72 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
MMLU-by-task - Abstract Algebra34Accuracy (%)82.1
MMLU-by-task - High School Mathematics30.74Accuracy (%)81.5
MMLU-by-task - College Mathematics35Accuracy (%)75.2
MMLU-by-task - Elementary Mathematics31.75Accuracy (%)65.6
LIBRA - ru2WikiMultihopQA35.16Dataset Total Score (%)64.3
MMLU-by-task - Machine Learning34.82Accuracy (%)60.6
LIBRA - ruQasper4.97Dataset Total Score (%)57.1
MMLU-by-task - High School European History63.03Accuracy (%)56.4
MMLU-by-task - High School Statistics39.81Accuracy (%)54.6
MMLU-by-task - TruthfulQA MC129.38Accuracy (%)54.2
MMLU-by-task - High School US History62.25Accuracy (%)54
MMLU-by-task - High School Geography57.58Accuracy (%)52.8

Interactive version: theaggregate.ai/model?slug=longchat-7b-v1-5-32k · How the rankings work · Data refreshed daily, snapshot 2026-07-22.