Llama 3.1
Meta·
LLMsopen-weightlocalMulti-GPU Self-Host
Overview
Massive 405B open-source model competing with frontier closed models.
Capabilities and innovations
405B Parameters128k ContextTool useMultilingual405B Open Weights128k ContextSynthetic Data Scaling
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 46.7% | unsourced |
| MMLU | 88.6% | unsourced |
- Humanity’s Last Exam (no tools): not reported by the vendor
- SWE-bench Verified: not reported by the vendor
- SWE-bench Pro: benchmark did not exist at release
- MMLU-Pro: benchmark did not exist at release
- Terminal-Bench 2.x: benchmark did not exist at release
- MMMU-Pro: benchmark did not exist at release
Architecture and hardware
- Parameters
- 405B total (Dense)
- Estimated VRAM at Q4
- ~233 GB, Multi-GPU Self-Host class
Links
More from Meta
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.