Blog & Demos

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

How SLMs Can Enable On-Device RAG - Making Industrial Machinery More Usable
GuideQuestion AnsweringOn-Prem / Edge

How SLMs Can Enable On-Device RAG - Making Industrial Machinery More Usable

Fine-tuned 1B parameter models can match the accuracy of 3B base models on domain-specific documentation, making on-device RAG viable for industrial equipment without expensive AI-optimized hardware. We tested this on a Siemens PLC manual and achieved a +16 percentage point accuracy gain through distillation.

Making FunctionGemma Work: Multi-Turn Tool Calling at 270M Parameters
DemoTool Calling

Making FunctionGemma Work: Multi-Turn Tool Calling at 270M Parameters

Google's FunctionGemma scores just 10-39% on multi-turn tool calling out of the box, but after fine-tuning with distil labs it reaches 90-97% accuracy across three benchmarks, matching or exceeding a 120B teacher model at 270M parameters.

When Does Reinforcement Learning Help Small Language Models?
BenchmarkClassificationQuestion AnsweringTool Calling

When Does Reinforcement Learning Help Small Language Models?

A controlled experiment across 12 datasets reveals that adding RLVR after SFT consistently improves text generation tasks (+2.0pp) but provides no reliable benefit for structured tasks like classification and function calling.

The LLM in Your Voice Assistant Is the Latency Bottleneck. Replace It with an SLM.
GuideTool CallingOn-Prem / Edge

The LLM in Your Voice Assistant Is the Latency Bottleneck. Replace It with an SLM.

Voice assistants on cloud LLMs are slow and expensive per turn. A fine-tuned SLM is cheaper and faster per request with equal-or-better accuracy on bounded tasks: brain-stage latency drops from ~700ms to ~40ms, and per-turn cost from cloud-API rates to server-amortized pennies.

pytest-generator: AI-Powered Unit Test Generation
DemoOn-Prem / Edge

pytest-generator: AI-Powered Unit Test Generation

Generate high-quality pytest test cases from Python function signatures and docstrings. Runs entirely on your local machine with zero API costs and complete privacy.

Helping Rocketgraph's customers with an OpenCypher-specialized small language model
Case StudyOn-Prem / EdgeTool Calling

Helping Rocketgraph's customers with an OpenCypher-specialized small language model

How distil labs partnered with Rocketgraph to finetune a small language model specialized in translating user questions to Rocketgraph-compliant Cypher queries on IBM Power hardware.

Teaching Small Language Models New Skills - Training a Local Cybersecurity Agent
Case StudyAgentic AIOn-Prem / Edge

Teaching Small Language Models New Skills - Training a Local Cybersecurity Agent

How distil labs partnered with Octodet to train a small language model that outperforms LLMs 30x its size at analyzing cybersecurity logs, while running entirely on-premises to meet strict privacy requirements.

Vibe-Tuning: The Art of Fine-Tuning Small Language Models with a Prompt
GuideClassification

Vibe-Tuning: The Art of Fine-Tuning Small Language Models with a Prompt

Fine-tuning is a pain – you need datasets, ML expertise, and a stack of GPUs just to get started. Not anymore. With model vibe-tuning, you go from prompt to production-ready model without these headaches. This blog post shows you exactly how to build one, starting with just a prompt.

AI Slop Detector: Catch AI-generated text with a 270M model that runs in your browser
DemoClassificationOn-Prem / Edge

AI Slop Detector: Catch AI-generated text with a 270M model that runs in your browser

A fine-tuned 270M parameter model that detects AI-generated text entirely in your browser: no API keys, no cloud, no data leakage. Matches 120B teacher accuracy at 400x smaller size.