ChatCrystal FAQs

Question 1

What LLM and embedding providers does ChatCrystal support?

Accepted Answer

ChatCrystal offers extensive multi-provider support for LLMs and embeddings, including Ollama (for local inference), OpenAI, Anthropic, Google AI, and Azure OpenAI. It's also compatible with any OpenAI-compatible API, providing flexibility in choosing your preferred models and services.

Question 2

What is ChatCrystal and how does it help developers?

Accepted Answer

ChatCrystal is a local-first tool designed to crystallize knowledge from your AI coding conversations. It helps developers by importing, summarizing, and semantically searching their interactions with AI assistants, building a personal, searchable knowledge base to prevent losing valuable solutions and insights over time.

Question 3

Which AI coding tools can ChatCrystal integrate with?

Accepted Answer

ChatCrystal offers multi-source ingestion with real-time file watching for popular AI coding tools such as Claude Code, Cursor, Codex CLI, Trae, and GitHub Copilot. This allows you to centralize and manage conversations from various development environments.

Question 4

How does ChatCrystal ensure summarized knowledge is accurate and searchable?

Accepted Answer

It uses structured LLM summarization with Zod schema for guaranteed valid output and automatic retries, ensuring high-quality notes. For retrieval, it leverages embedding-powered semantic search with relation-aware result expansion, allowing you to find precise and contextually relevant information.

Question 5

Is ChatCrystal cloud-based or does it run locally?

Accepted Answer

ChatCrystal runs entirely on your local machine, either as a desktop application (Electron) or a local web server. This ensures your conversation data remains private and under your full control, with no dependencies on cloud services or recurring subscriptions.

ChatCrystal

ChatCrystal

主な機能

ユースケース

主な機能

ユースケース