LLaMA 3.1 vs o1-preview
Meta LLaMA 3.1 | OpenAI o1-preview | |
|---|---|---|
| Overview | ||
| Company | Meta | OpenAI |
| Release date | Jul 23 2024 | Sep 12 2024 |
| Access | Open Weight | Proprietary |
| Benchmarks | ||
Nonsense detection BullshitBench v2Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 14% | — |
| Timeline | ||
| Release gap | LLaMA 3.1 shipped 51 days before o1-preview | |
Which is better: LLaMA 3.1 or o1-preview?
LLaMA 3.1 and o1-preview don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. LLaMA 3.1 shipped 51 days before o1-preview, so benchmark comparisons should account for the intervening progress.
LLaMA 3.1 is open weight, while o1-preview is proprietary.
Direct benchmark comparisons are unavailable — LLaMA 3.1 and o1-preview don't publish scores on any of the same benchmarks.
Frequently asked questions
LLaMA 3.1 was released by Meta on Jul 23 2024.
o1-preview was released by OpenAI on Sep 12 2024.
LLaMA 3.1 is an open weight model released by Meta. o1-preview is a proprietary model released by OpenAI.
Other comparisons
LLaMA 3.1 vs Claude Sonnet 5o1-preview vs Claude Sonnet 5LLaMA 3.1 vs Gemini Omnio1-preview vs Gemini OmniLLaMA 3.1 vs Grok 4.5o1-preview vs Grok 4.5LLaMA 3.1 vs DeepSeek-V4-Proo1-preview vs DeepSeek-V4-ProLLaMA 3.1 vs Mistral Medium 3.5o1-preview vs Mistral Medium 3.5LLaMA 3.1 vs Kimi K2.7 Codeo1-preview vs Kimi K2.7 Code