Compare LLM Models
Select 2-5 models to compare side by side.
Select models to compare
0 / 5 selected
Top downloads
Select 2-5 models
| Attribute |
meta-llama
|
|---|---|
| Company | meta-llama |
| Release Date | 2025-04-05 |
| Parameters (B) | 400 |
| Architecture | MoE Transformer |
| Context Window | 1048576 |
| Input Modalities | text, image |
| Open Source | Yes |
| License | other |
| Score | 14 |
| VRAM (GB) | 640 |
| Compute | 8x NVIDIA H100 80GB (FP8 quantized on a single DGX host); BF16 weights require multi-node deployment |
| Benchmark: HLE | 5.7 |
| Benchmark: MATH | 61.2 |
| Benchmark: MBPP | 77.6 |
| Benchmark: MMLU-Pro | 80.5 |
| Benchmark: SWE-Bench-Pro | 5.2 |
| Pricing (Input) | — |
| Pricing (Output) | — |