LLaMA 3.1 vs o3
Meta LLaMA 3.1 | OpenAI o3 | |
|---|---|---|
| Overview | ||
| Company | Meta | OpenAI |
| Release date | Jul 23 2024 | Apr 16 2025 |
| 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% | 26%Best |
| Timeline | ||
| Release gap | LLaMA 3.1 shipped 267 days before o3 | |
Which is better: LLaMA 3.1 or o3?
o3 leads LLaMA 3.1 on 1 of the 1 benchmark they both report (BullshitBench v2). LLaMA 3.1 shipped 267 days before o3, so benchmark comparisons should account for the intervening progress.
LLaMA 3.1 is open weight, while o3 is proprietary.
On BullshitBench v2, o3 leads at 26% vs LLaMA 3.1 at 14%.
Frequently asked questions
LLaMA 3.1 was released by Meta on Jul 23 2024.
o3 was released by OpenAI on Apr 16 2025.
LLaMA 3.1 is an open weight model released by Meta. o3 is a proprietary model released by OpenAI.