A compact, conversation-focused model that handles everyday dialogue and basic reasoning with reasonable fluency. It follows instructions reliably within familiar territory but shows its limits quickly on complex multi-step reasoning or specialized knowledge. Think of it as a capable junior assistant — dependable for straightforward tasks, but not the one you'd send to tackle nuanced or technically demanding problems.
| Benchmark | Score | Type | Recorded |
|---|---|---|---|
| IFEval | 39.9 | accuracy | 2d ago |