Kiln

What it does

Kiln is a desktop workbench and optional Python library for teams developing AI systems. Engineers and data scientists can work with datasets, prompts, models, and documents; build retrieval-augmented generation (RAG) tasks; and evaluate changes with LLM-as-judge scoring, golden datasets, or human ratings. It also supports prompt optimization, fine-tuning, synthetic-data generation, and composing tools and sub-agents. The AI assistant can help create experiments and optimize systems.

The desktop app is offered for macOS, Windows, and Linux. Developers can use the Python library from a codebase or notebook, or choose the documented self-hostable REST API. Product managers, QA staff, and subject-matter experts can contribute feedback and ratings. The Individual tier is free; Team access is requested, and Enterprise pricing is custom. The Python library is MIT-licensed, while the desktop app is source-available. Kiln says datasets and evals live on the user's machine and can sync through a Git repository; Kiln Pro features run on Kiln's servers. Pricing does not specify external model or compute charges.

What it can do

• Build RAG document and search tasks
• Create and run evaluations, including LLM-as-judge scoring
• Optimize prompts against evaluations
• Fine-tune models and generate synthetic data
• Compose tools and sub-agents
• Sync datasets with Git and collect team feedback

Works with

• Git
• MCP

Before you choose

• Kiln Pro features run on Kiln's servers; the pricing page says datasets and evals live on the user's machine and sync through a repository the user controls.
• The desktop app is source-available, not described as open source; the Python library is MIT-licensed.
• Team access is request-based and Enterprise pricing is custom. External model-provider or compute charges are not specified.

Listed under

• Coding
• AI development
• Developers
• Evals
• Fine-tuning
• Open source
• Prompt optimization

Get started

Kiln offers a free download; Team access is requested separately.

1. Download the desktop app for a listed operating system, or install the Python library with `pip install kiln_ai`.
2. For library use, load a Kiln project or dataset into your notebook or codebase; the library is optional for app users.
3. Select the app, library, or self-hostable REST API route; request access for the Team tier.

Individual is free. Team requires access; Enterprise is custom, with annual contracts mentioned. Kiln Pro features run on Kiln's servers. External model-provider or compute charges are not specified.

Source: kiln.tech/pricing, checked 1 Nov 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 1 Nov 2026. Each source supports part of the overview; availability and prices can change.

• Kiln AI - Build, Evaluate, and Optimize AI Systems ↗kiln.tech
• Python Library Setup | Kiln AI Docs ↗docs.kiln.tech
• Pricing—Kiln ↗kiln.tech
• About Kiln AI - Why We Built Kiln, and Where We're Heading ↗kiln.tech
• Contact Sales - Kiln AI ↗kiln.tech
• Kiln AI - Build, Evaluate, and Optimize AI Systems ↗kiln.tech

Company website

kiln.tech: product, pricing and policy pages

YC directory mirror

Batch and company name

Human review

Checked by a person

knowledge_base

manual_review

Last checked 1 Nov 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.