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 |
ibm-granite
|
|---|---|
| Company | ibm-granite |
| Release Date | 2026-07-07 |
| Parameters (B) | 3 |
| Architecture | MoE Transformer |
| Context Window | 8192 |
| Input Modalities | text |
| Open Source | Yes |
| License | apache-2.0 |
| Score | 0 |
| VRAM (GB) | 8 |
| Compute | NVIDIA GPU with 8GB+ VRAM for BF16 inference; Flash Attention 3 or 4 recommended (SDPA not supported due to attention sinks) |
| Benchmark: GSM8K | 39.6 |
| Benchmark: MMLU | 54.1 |
| Pricing (Input) | — |
| Pricing (Output) | — |