Llama 3.3
Meta·
LLMsopen-weightlocalWorkstation
Overview
Efficient 70B text model matching Llama 3.1 405B on key benchmarks.
Capabilities and innovations
70B Parameters405B-level performanceCost efficientEfficiency BreakthroughDistillation Techniques
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 46.7% | unsourced |
| MMLU | 86.3% | 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
- 70B total (Dense)
- Estimated VRAM at Q4
- ~41 GB, Workstation class
Links
More from Meta
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.