Blog & Demos

Tutorials, case studies, benchmarks, and open-source demos — everything you need to build with small language models.

SHELLper: Multi-turn Bash Function Calling Model
DemoTool CallingOn-Prem / EdgeAgentic AI

SHELLper: Multi-turn Bash Function Calling Model

A fine-tuned 0.6B model that converts natural language into bash commands with 100% multi-turn tool call accuracy. Runs locally with full privacy.

Text2SQirreL 🐿️: Query your data in plain English
DemoQuestion AnsweringOn-Prem / Edge

Text2SQirreL 🐿️: Query your data in plain English

A fine-tuned small language model that converts plain English questions into executable SQL queries. Runs locally with no API keys, no cloud dependencies, and full privacy — matching the accuracy of a 685B teacher model.

We Benchmarked 12 Small Language Models Across 8 Tasks to Find the Best Base Model for Fine-Tuning
BenchmarkClassificationQuestion Answering

We Benchmarked 12 Small Language Models Across 8 Tasks to Find the Best Base Model for Fine-Tuning

A systematic benchmark of 12 small language models across 8 tasks reveals Qwen3-4B as the best for fine-tuning, with fine-tuned models matching or exceeding a 120B+ teacher. Smaller models like Llama-3.2-1B show the highest tunability.

Gitara: How we trained a 3B Function-Calling Git Agent for Local Use
DemoTool CallingAgentic AI

Gitara: How we trained a 3B Function-Calling Git Agent for Local Use

We fine-tuned a small, tool-calling language model to turn plain-English questions into git commands with the accuracy of a cloud LLM.

Resume Roaster AI: Brutally Honest Resume Critique with a Local SLM
DemoInformation ExtractionOn-Prem / Edge

Resume Roaster AI: Brutally Honest Resume Critique with a Local SLM

A fine-tuned Llama-3.2-3B model that generates sarcastic resume critiques and professional improvement suggestions. Runs entirely locally to keep your personal data private.

distil-commit-bot: AI-Powered Commit Messages for TypeScript
DemoOn-Prem / Edge

distil-commit-bot: AI-Powered Commit Messages for TypeScript

A fine-tuned 0.6B parameter SLM that generates commit messages for TypeScript codebases. Runs locally via Ollama, achieving 90% accuracy compared to a 120B teacher model — at 200x smaller size.

distil-localdoc.py: Automatic Python Documentation Generation
DemoOn-Prem / Edge

distil-localdoc.py: Automatic Python Documentation Generation

A fine-tuned Qwen3 0.6B model that generates complete, properly formatted Google-style docstrings for your Python code — runs locally to keep your proprietary code secure.

distil-expenses: Local Personal Expense Summaries with SLMs
DemoTool CallingQuestion AnsweringOn-Prem / Edge

distil-expenses: Local Personal Expense Summaries with SLMs

Fine-tuned Llama 3.2 models (1B and 3B) for personal expense analysis. Runs locally via Ollama — query your spending data with natural language while keeping your financial data completely private.

distil NPC: Small Language Models for Video Game Characters
DemoQuestion AnsweringOn-Prem / Edge

distil NPC: Small Language Models for Video Game Characters

A family of small language models specialised for conversational NPCs in video games. Enables natural language interaction with game characters running entirely on-device, no network required.