Llama 4 Scout
Announced on meta.com
Benchmark results
| Benchmark | Score | Source check |
|---|---|---|
| MMLU-Pro0-shot | 74.3% | |
| LiveCodeBenchproblems from Oct 2024 to Feb 2025; 0-shot; averaged over multiple generations | 32.8% |
Sources
Reported by Meta
- The Llama 4 herd: The beginning of a new era of natively multimodal AI innovationai.meta.com ยท 5 Apr 2025
Original image
