AI Gist is committed to maximizing the value of AI prompt words collected by individuals.
Supports core functions such as variable replacement, Jinja template, AI generation and tuning, historical version recording, and cloud backup.
project information
- AI Gist Is a “privacy-first” AI prompt word management tool that focuses on helping users efficiently manage and utilize their own prompt word collection.
- The open source agreement is AGPL-3.0。
- Support multiple platforms (Windows/macOS/Linux) and multiple languages (Simplified Chinese, Traditional Chinese, English, Japanese).
main function
Prompt word template management
- support the use of Jinja templates, you can insert variables (such as
{{variable name}}), only need to fill in the specific value when using it to generate content. - Multi-view management: Provide card views, table views, classification views, etc. to facilitate viewing and organizing prompt words.
- Support filtering functions such as labeling, classification, scoring, and collection to facilitate rapid search and classification management.
- Complete historical record function: It is convenient to trace back, reuse and continuously optimize prompt words.
AI helps optimization of prompt words
- Support access to multiple AI models, including local models (Ollama, LM Studio) and online models (such as OpenAI, DeepSeek).
- Has the following AI functions:
- AI generates prompt words: Quickly generate prompt word templates directly from AI.
- AI tuning prompt words: Make existing prompt words more specific and richer.
- AI automatically extracts variables: AI identifies and extracts variable positions without manual hollowing out.
Privacy and data security
- local priority storage: By default, all data is stored locally, without networking to ensure user privacy and security.
- Provide data export and import functions (support CSV format), and users have full data control.
- Supports cloud backup and recovery, including WebDAV and iCloud, for easy synchronization among multiple devices.
Latest update highlights
According to information posted on SourceForge:
- Version 0.14.0 (released on 2025-06-27): Reconstructed the synchronization functions of WebDAV and iCloud to simplify logic, enhance the stability of backup and recovery, and avoid real-time synchronization conflicts.
- Version 0.12.0 (released on 2025-06-23): Optimize WebDAV synchronization logic and directory structure to achieve true one-click synchronization.
Summary overview
| aspects | function description |
|---|---|
| template management | Use Jinja’s template with variables, support multiple views and filtering mechanisms, and have history function |
| AI helps | Support the generation and refinement of prompt words and automatic extraction of variables, and connect to multiple AI models |
| Privacy Protection | Default local storage, support import, export and cloud backup (WebDAV / iCloud) |
| Cross-platform and multi-language | Compatible with Windows/macOS/Linux, multi-language interface |
| stable synchronization | The latest version improves the synchronization mechanism of WebDAV and iCloud to enhance stability |
Github：https://github.com/yarin-zhang/AI-Gist
Oil tubing: