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 |
openbmb
|
|---|---|
| Company | openbmb |
| Release Date | 2026-05-23 |
| Parameters (B) | 3 |
| Architecture | Llama Transformer (GQA) |
| Context Window | 32768 |
| Input Modalities | text |
| Open Source | Yes |
| License | apache-2.0 |
| Score | 0 |
| VRAM (GB) | 16 |
| Compute | Single NVIDIA GPU with 16GB+ VRAM (or Huawei Ascend 910B NPU for native ternary training); DeepSpeed or HuggingFace Trainer with trust_remote_code=True |
| Benchmark: ARC | 79.0 |
| Benchmark: BBH | 68.3 |
| Benchmark: C-Eval | 75.9 |
| Benchmark: GSM8K | 79.5 |
| Benchmark: MMLU | 64.4 |
| Pricing (Input) | — |
| Pricing (Output) | — |