Kiln

Build, Evaluate, and Optimize AI Systems. Includes evals, RAG, agents, fine-tuning, synthetic data generation, dataset management, MCP, and more.

What is Kiln?

Build, Evaluate, and Optimize AI Systems. Includes evals, RAG, agents, fine-tuning, synthetic data generation, dataset management, MCP, and more.

Overview of Kiln

What it is

Kiln is a free desktop app and open-source Python library for building, evaluating, and optimizing AI systems. It covers the full AI development loop: evals, prompt optimization, RAG, agents, fine-tuning, synthetic data generation, dataset management, and MCP. The app runs locally, supports bring-your-own API keys or fully offline use with Ollama, and syncs via Git for team collaboration. The Python library is MIT-licensed and can be used to deploy the same tasks to production.

Key features

  • Eval Builder: Auto-generates evals (judge + synthetic eval dataset) and aligns to user preference in about 10 minutes.
  • Auto-Optimize: Automatically finds the best way to run an AI task, optimizing prompt, model selection, tools, skills, subagents, parameters, and more.
  • AI Assistant: An AI data-science partner that proposes improvements, optimizes prompts, runs experiments, and creates evals.
  • RAG: Drag-and-drop documents (PDF, image, video, audio) to create a RAG, with auto-generated RAG evals from the documents.
  • Subagents: Compose multi-agent hierarchies, each running in its own focused context window.
  • Fine-Tuning: Zero-code fine-tuning across 60+ models (Qwen, Llama, GPT, Gemini, etc.) on Fireworks, Together, and Vertex, with serverless deployment included.
  • Synthetic Data Generation: Generate data for evals or fine-tuning in minutes.

Use cases

Kiln is suitable for teams building AI products who need to iterate, evaluate, and optimize across multiple techniques. Use cases include: evaluating and improving prompt quality, building RAG systems from documents, composing multi-agent workflows, fine-tuning models without writing code, generating synthetic datasets for training or evaluation, and collaborating across roles (PMs, subject-matter experts, QA) without requiring coding skills. The library allows engineers to deploy the same tasks in production.

Getting started

Download Kiln Desktop for macOS, Windows, or Linux, then follow the 5-minute quickstart to run your first task. For the Python library, install via pip:bashpip install kiln-ai

License and pricing

Kiln's core Python library and REST server are MIT-licensed. The desktop app is source-available, free to use, and built on the fair-code model. Datasets are open JSON and owned by the user. Kiln Pro is an opt-in service that adds the AI Assistant, Auto-Optimize, and the Eval Builder; the core app remains fully functional without it. The Kiln name and logos are trademarks of Chesterfield Laboratories Inc.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/kiln-aikiln.svg)](https://yunodb.com/tools/kiln-aikiln?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/kiln-aikiln?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/kiln-aikiln.svg" alt="Featured on YunoDB" height="20"></a>