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 |
google
|
|---|---|
| Company | |
| Release Date | 2025-03-12 |
| Parameters (B) | 27 |
| Architecture | Decoder-only Transformer |
| Context Window | 128000 |
| Input Modalities | text, image |
| Open Source | Yes |
| License | gemma |
| Score | 0 |
| VRAM (GB) | 54.8 |
| Compute | Single NVIDIA A100 80GB for BF16 inference (~55GB); ~27GB with INT8 or ~14GB with INT4 quantization on consumer GPUs |
| Benchmark: BBH | 87.6 |
| Benchmark: DROP | 77.2 |
| Benchmark: GPQA | 36.8 |
| Benchmark: GSM8K | 95.9 |
| Benchmark: HellaSwag | 85.6 |
| Benchmark: MATH | 89 |
| Benchmark: MBPP | 74.4 |
| Benchmark: SWE-Bench-Pro | 11.4 |
| Pricing (Input) | 0.08 |
| Pricing (Output) | 0.45 |