Thanks to @Masterchiefm (https://github.com/Masterchiefm/pymol-ai-assistant). This is a description of fixes and added features.

The following features have been added:

UI text switching: Now automatically detects the user's language environment, no need to manually press buttons to switch between Chinese and English

KIMI API online search capability: Prevents offline models from giving incorrect answers that affect experiments

Adjusted UI image upload position: Fixed some high-DPI screen bugs, adjusted UI fonts and the logic for displaying advanced configuration in the config folder

Custom prompt adjustment window: Set up using the original log location

Thinking mode: Thinking content is collapsed by default when using thinking mode

Control PyMOL with natural language. Make molecular visualization simple and efficient.

简体中文 | English (Current)

As shown in the screenshot, you only need to describe your needs in everyday language, and the AI will directly control PyMOL to complete complex molecular visualization tasks.

---

✨ Features

Feature	Description
🤖 AI Chat	Control PyMOL with natural language, no need to memorize complex commands
🌊 Streaming	Real-time display of AI thinking process and output with color distinction
🔧 Tool Calling	AI can directly operate PyMOL: load structures, set styles, save images, etc.
⚙️ Config Management	Support multiple API configs (SiliconFlow, OpenAI, etc.), import/export supported
📋 Logging	Record all conversations and tool calls, with filtering and export
🌐 Bilingual	One-click switching between Chinese/English interface, language preference auto-saved
📜 Chat History	View complete chat_history JSON for debugging and analysis
📦 Auto Dependencies	Automatically check and install required dependencies on installation

---

📥 Installation

Install via Plugin Manager

1. Download Plugin

   • Download PyMOL-AI-Agent-Plugin.zip from release page

2. Installation Steps

   

   • Open PyMOL → Plugin → Plugin Manager
   • Click "Install New Plugin"
   • Select the downloaded zip file
   • Restart PyMOL

---

🚀 Usage

Quick Start

1. Launch PyMOL

2. Menu bar: Plugin → AI Assistant

3. Configure API for first-time use:

   

   • Click "⚙️ Config" button (or "⚙️ 配置" in Chinese interface)
   • Add API configuration (URL, Key, Model)
   • Supports SiliconFlow, OpenAI, and other OpenAI API-compatible services

4. (Optional) Click "🌐 中文" to switch to Chinese interface. Language preference is automatically saved.

API Configuration Example

SiliconFlow (Recommended for China users)

API URL: https://api.siliconflow.cn/v1
Recommended Models:
  - Pro/moonshotai/Kimi-K2.5  # Best overall, paid
  - Pro/zai-org/GLM-4.7       # Paid
  - deepseek-ai/DeepSeek-R1 # Free model

Registration Bonus:

• Get 16 CNY voucher with invite link: https://cloud.siliconflow.cn/i/Su2ao83G
• Or visit Kimi official website for monthly plans

After configuration, type commands in the input box and press Enter to send.

---

🖥️ Interface Guide

Tab Layout

Tab	Content
💬 Chat	Chat interaction interface
📋 Logs	System logs and debug information
📜 History	View complete chat_history JSON

Message Styles

• 👤 User Message: Blue background, displayed on right
• 🤖 AI Message: Green background, displayed on left
  • 💭 Thinking Process: Gray italic
  • Formal Output: Normal text
  • ⚙️ Tool Call: Orange display
  • ✓ Tool Result: Green/Red status

---

💡 Example Commands

Example 1: Load and Visualize Molecule

Load PDB 1ake, select each chain and color them, display surface with transparency

Example 2: Measure Distance

What is the distance between the two residues I selected? Show me the measurement.

More Examples

Rotate view 90 degrees, then save as image

Create a selection of all atoms in chain A

Execute PyMOL script /path/to/script.pml

Run command: load 1ake; show cartoon; color chain

Load and run Python script /path/to/setup.py

---

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

🤝 Contributing

Issues and Pull Requests are welcome!

---

☕ Support

If this project helps you, please consider giving it a Star ⭐!

---

📝 Special Statement

This project was primarily developed with the assistance of AI models:

• Kimi K2.5 (Moonshot AI) - Core architecture and functionality
• GLM-4.7 (Zhipu AI) - Code optimization and feature enhancement

With the help of AI, the developer realized this powerful PyMOL intelligent plugin by describing requirements in natural language.

---

Made with ❤️ and AI (Kimi K2.5 & GLM-4.7)