o1 benchmarks and performance, in charts
Every chart we tracked that shows o1, from the lab itself and from the trackers people trust, with how far each one travelled.
o1 on its own benchmark table
Drawn from the numbers openai.com published as text (OpenAI o1 System Card | OpenAI, read Dec 5, 2024). Each panel is one test; the highlighted bar is o1. Source: openai.com.
See the numbers
| harassment | o1-preview 0.995, o1 0.99, GPT-4o 0.985, o1-mini 0.975 |
|---|---|
| harassment/threatening | GPT-4o 0.995, o1 0.995, o1-preview 0.99, o1-mini 0.99 |
| hate | o1-preview 0.975, o1 0.965, o1-mini 0.96, GPT-4o 0.945 |
| hate/threatening | o1-mini 0.985, o1-preview 0.98, o1 0.98, GPT-4o 0.975 |
| self-harm | o1 0.965, o1-preview 0.955, o1-mini 0.955, GPT-4o 0.945 |
| self-harm/instructions | o1 0.995, o1-mini 0.99, o1-preview 0.985, GPT-4o 0.95 |
Official charts from OpenAI
As published on X, Hugging Face, GitHub and their blog. Each card opens our own page for the chart.
From benchmark trackers and others
More charts that feature o1
Benchmark trackers, other labs and researchers. Each card opens our own page for the chart.
Make it yours
Your numbers, in a look-alike of the original.
Starts from the real numbers of "Accuracy / Percentile (%)" by deepseek. Replace every number with yours before you post. Make one like this, step by step
Look-alike score 8.3 of 10. Free pictures carry a small "Made with ViralChart" line. That line is how other people find us.
Questions people ask
How does o1 perform?
We hold 3 charts that show o1, posted between Jan 2025 and Aug 2025, 0 of them from OpenAI. The most seen is "AI model benchmark comparison" by Kimi (Moonshot AI) with 420,401 views.
o1 vs GPT-4o: which is better?
2 charts put them side by side. Which wins depends on the benchmark; each chart shows its own test and who published it.
Who publishes o1 benchmark charts?
Kimi (Moonshot AI), Arena, Epoch AI, OpenAI, @ArtificialAnlys, @SpaceXSI, @EpochAIResearch, @METR_Evals, @arena, zai-org/GLM-4