Who leads the HellaSwag benchmark?

gpt-4o currently leads HellaSwag with a score of 95.3 on accuracy.

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

The state-of-the-art result on HellaSwag is 95.3 (accuracy), achieved by gpt-4o as of 2026.

How many models are tracked on HellaSwag?

Codesota tracks 4 models on HellaSwag.

When was the HellaSwag leaderboard last updated?

The HellaSwag leaderboard on Codesota includes results through 2026.

Codesota · Benchmark · HellaSwagHome/Leaderboards/HellaSwag

Unknown

HellaSwag.

Name: HellaSwag Benchmark Results
Creator: Unknown
Published: 2026-01-01
License: https://creativecommons.org/licenses/by/4.0/

70K sentence completion problems testing commonsense natural language inference.

Paper ↗Leaderboard ↓

§ 01 · 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 HellaSwag. 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

Muted rows were not state of the art when published — an earlier or same-year result already scored better.

Rank	Model	Trust	Score	Year	Links	Fix
01	gpt-4o Non-API entry from src	unverified	95.3	2026	N/A	Looks wrong?
02	gemini-15-pro Non-API entry from src	unverified	92.5	2026	N/A	Looks wrong?
03	claude-35-sonnet Non-API entry from src	unverified	89	2026	N/A	Looks wrong?
04	llama-3-70b Non-API entry from src	unverified	88	2026	N/A	Looks wrong?

§ 04 · Submit a result

Add to the leaderboard.

Submit a Result

← Back to Leaderboards