Llama 3.1 405B
- Made by
- Meta
- Released
- Announced in
- Introducing Llama 3.1: Our most capable models to dateai.meta.com
- Event
- Llama 3.1 405B
- API record
- /api/v1/models/llama-3-1-405b
Benchmark results
| Benchmark | Score | Source check |
|---|---|---|
| GPQA Diamond0-shot chain of thought | 51.1% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
| 0-shot; averaged over multiple generations · reported later, in The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation | 49.0% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
| MMLU-Pro5-shot chain of thought | 73.3% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
| 0-shot · reported later, in The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation | 73.4% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
| LiveCodeBenchproblems from Oct 2024 to Feb 2025; 0-shot; averaged over multiple generations · reported later, in The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation | 27.7% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
Sources
Reported by Meta
- Introducing Llama 3.1: Our most capable models to dateai.meta.com · 23 Jul 2024
Original image - The Llama 4 herd: The beginning of a new era of natively multimodal AI innovationai.meta.com · 5 Apr 2025
Original image
