Who leads the RACE benchmark?

Megatron-BERT currently leads RACE with a score of 90.9 on Accuracy.

What is the state-of-the-art score on RACE?

The state-of-the-art result on RACE is 90.9 (Accuracy), achieved by Megatron-BERT as of 2026.

How many models are tracked on RACE?

Codesota tracks 4 models on RACE.

When was the RACE leaderboard last updated?

The RACE leaderboard on Codesota includes results through 2026, with the earliest tracked result from 2019.

Codesota · Benchmark · RACEHome/Leaderboards/Language & Knowledge/Reading Comprehension/RACE

Unknown

RACE.

Name: RACE Benchmark Results
Creator: Unknown
Published: 2019-01-01
License: https://creativecommons.org/licenses/by/4.0/

Canonical multiple-choice reading comprehension benchmark built from English exams for Chinese middle and high school students. ~28K passages and ~100K questions. Evaluated as accuracy over RACE-M (middle) + RACE-H (high) combined.

Paper ↗Leaderboard ↓

§ 01 · SOTA history

Year over year.

§ 02 · Leaderboard

Results by metric.

Only 4 models on this benchmark

Help build the community leaderboard — submit your model results.

Found a wrong score or missing run?

Use row edits to send a sourced correction into moderation.

Add / edit result ↗Report issue ↗

Accuracy

Accuracy is the reported evaluation metric for RACE. Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.

Higher is better

Trust tiers for Accuracyverifiedpapervendorcommunityunverified

Rank	Model	Trust	Score	Year	Links	Edit
01	Megatron-BERT Megatron-BERT, combined RACE-M+H accuracy — seed — verify	paper	90.9	2026	Source ↗	Edit result
02	ALBERT (Ensemble) ALBERT ensemble, combined RACE-M+H accuracy — seed — verify	paper	89.4	2026	Source ↗	Edit result
03	ALBERT ensemble	unverified	89.4	2019	Paper ↗Code ↗	Edit result
04	RoBERTa	unverified	83.2	2019	Paper ↗Code ↗	Edit result

§ 04 · Submit a result

Add to the leaderboard.

← Back to Reading Comprehension