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 |
tiiuae
|
|---|---|
| Company | tiiuae |
| Release Date | 2026-01-12 |
| Parameters (B) | 0.1 |
| Architecture | Hybrid Transformer-Mamba |
| Context Window | 262144 |
| Input Modalities | text |
| Open Source | Yes |
| License | other |
| Score | 0 |
| VRAM (GB) | 0.5 |
| Compute | Runs on CPU or any modest GPU; ~0.18 GB for BF16 weights (~91M params), with GGUF quants as low as ~60–90 MB. Suitable for edge devices, Raspberry Pi, laptops, and integrated graphics. |
| Benchmark: ARC | 23.6 |
| Benchmark: BBH | 2.5 |
| Benchmark: GPQA | 0 |
| Benchmark: GSM8K | 19.2 |
| Benchmark: HellaSwag | 36.4 |
| Benchmark: HumanEval | 7.3 |
| Benchmark: IFEval | 66.1 |
| Benchmark: MATH | 2.0 |
| Benchmark: MBPP | 7.9 |
| Benchmark: MMLU | 27.3 |
| Benchmark: MMLU-Pro | 7.0 |
| Benchmark: TruthfulQA | 42.0 |
| Pricing (Input) | — |
| Pricing (Output) | — |