Atlas

Models

← All models

ChatGPT-4o Latest observed 2025-01-29

OpenAI · 2025-01-29

ChatGPT-4o Latest observed 2025-01-29 is a benchmark-entry representation of OpenAI's mutable `chatgpt-4o-latest` API alias. EQ-Bench rewrites the source identifier `openai/chatgpt-4o-latest` to this date-stamped label, which matches OpenAI's January 29, 2025 GPT-4o update in ChatGPT, but neither source identifies an immutable dated API checkpoint.

Benchmark scores

BenchmarkScore
Confabulations Leaderboard Confabulation Rate26.7
Confabulations Leaderboard Non-Response Rate6.5
Confabulations Leaderboard Weighted Score16.6
Creative Writing v3 Elo1352
Creative Writing v3 Rubric Score15.4
Elimination Game TrueSkill Mu3.5
Fiction.liveBench66.7
PGG-Bench TrueSkill Mu9.9
Step Game TrueSkill Mu1.1
Loading Atlas data…