A compact, instruction-following model that punches reasonably well for its size. It handles everyday tasks like summarization, Q&A, and light coding with decent fluency, though it can lose coherence on complex multi-step reasoning or nuanced topics. Its open-weight nature means it runs locally on modest hardware, making it a practical workhorse rather than a frontier performer.
| Benchmark | Score | Type | Recorded |
|---|---|---|---|
| IFEval | 74.1 | accuracy | 2d ago |