We build efficient models that help AI assistants decide which tool to use鈥攐r when no tool is needed.
Our Lapis SAL family focuses on English鈥揝panish tool routing, with a larger teacher, a distilled student, and an MLX release for local inference on Apple Silicon.
Assistants often need to choose between searching the web, checking a calendar, sending a message, or answering directly.
A dedicated routing model can handle this decision before the request reaches a larger language model or an external service. This supports local processing and gives developers explicit control over the available tools.
| Model | Purpose |
|---|---|
| Lapis SAL 150M | Distilled PyTorch student for research, further training, and export. |
| Lapis SAL 421M | Larger PyTorch teacher used in the distillation workflow. |
| Lapis SAL 150M MLX | Apple Silicon release with FP16 inference and optional INT8 quantization at load time. |
The student鈥檚 150M name identifies its encoder family. The complete routing model contains approximately 164M learned parameters.
Calculator 路 Calendar 路 Clock 路 Code 路 Contacts 路 Health 路 Location 路 Mail 路 Maps 路 Messages 路 Reminders 路 Weather 路 Web Search
A NONE option covers requests that do not require an available tool. Developers supply the tool descriptions at inference time.
The student learns from a larger teacher through knowledge distillation. The workflow includes supervised adaptation, calibration, and deployment export.
On our private, synthetic, held-out benchmark of 1,000 English鈥揝panish examples:
These results describe this benchmark and do not guarantee performance on other datasets or real-world traffic. See each model card for evaluation details and limitations.
Choose a model above and follow its README for installation and usage.
These releases use a custom routing runtime. They select tools; they do not execute them, generate conversational responses, or provide medical advice.
English & Spanish 路 Knowledge distillation 路 Local inference 路 Apache 2.0