Godot AI聊天
AI Skill Hub 推荐使用：Godot AI聊天 是一款优质的AI工具。AI 综合评分 7.5 分，在同类工具中表现稳健。如果你正在寻找可靠的AI工具解决方案，这是一个值得深入了解的选择。
📚 深度解析
**为什么要使用开源工具而非商业 SaaS？**
对于个人开发者和有隐私需求的用户，本地部署的开源工具意味着数据不离本机，不受第三方服务商的数据政策约束。同时，开源工具通常没有使用次数限制和月度费用，一次安装即可长期使用，对于高频使用场景的总拥有成本（TCO）远低于订阅制商业工具。
**安装与环境准备**
Godot AI聊天 依赖 GDScript 运行环境。建议通过 pyenv（Python）或 nvm（Node.js）管理 GDScript 版本，避免全局环境污染。对于新手用户，推荐先创建虚拟环境（python -m venv venv && source venv/bin/activate），再安装依赖，这样即使出现问题也可以随时删除虚拟环境重新开始，不影响系统稳定性。
**社区与维护**
GitHub Issue 和 Discussion 是获取帮助的最快渠道。在提问前建议先检查 Closed Issues（已关闭的问题），大多数常见问题都已有解答。遇到 Bug 时，提供 pip list 的输出、完整错误堆栈和最小可复现示例，能显著提高开发者响应速度。AI Skill Hub 将持续追踪 Godot AI聊天 的版本更新，及时通知重要功能变化。
📋 工具概览
Godot AI聊天 是一款基于 GDScript 开发的开源工具，专注于 Godot、AI、LLM 等核心功能。作为 GitHub 开源项目，它拥有活跃的社区支持和持续的版本迭代，代码完全透明可审计，支持本地部署以保护数据隐私。无论是个人使用还是集成到企业工作流，都能提供稳定可靠的解决方案。
📖 中文文档
Godot AI聊天 是一款基于 GDScript 开发的开源工具，专注于 Godot、AI、LLM 等核心功能。作为 GitHub 开源项目，它拥有活跃的社区支持和持续的版本迭代，代码完全透明可审计，支持本地部署以保护数据隐私。无论是个人使用还是集成到企业工作流，都能提供稳定可靠的解决方案。
- 开源免费，支持本地部署，数据完全自主可控
- 活跃的 GitHub 开源社区，持续迭代更新
- 提供详细文档和使用示例，新手友好
- 支持自定义配置，灵活适配不同使用环境
- 可作为基础组件集成进现有技术栈或进行二次开发
- 本地部署运行，保护数据隐私，满足合规要求
- 自定义集成到现有系统，扩展技术栈能力
- 作为开源基础组件进行商业化二次开发
# 克隆仓库 git clone https://github.com/snougo/Godot-AI-Chat cd Godot-AI-Chat # 查看安装说明 cat README.md # 按 README 完成环境依赖安装后即可使用
- 访问 GitHub 仓库页面
- 按照 README 文档完成依赖安装
- 根据系统环境完成初始化配置
- 参考官方示例或文档开始使用
- 遇到问题可在 GitHub Issues 中查找解答
# 查看帮助 godot-ai-chat --help # 基本运行 godot-ai-chat [options] <input> # 详细使用说明请查阅文档 # https://github.com/snougo/Godot-AI-Chat
# godot-ai-chat 配置说明 # 查看配置选项 godot-ai-chat --config-example > config.yml # 常见配置项 # output_dir: ./output # log_level: info # workers: 4 # 环境变量（覆盖配置文件） export GODOT_AI_CHAT_CONFIG="/path/to/config.yml"
Godot AI Chat
<img width="1624" height="1080" alt="截屏2026-03-02 22 49 06" src="https://github.com/user-attachments/assets/9fbb86cd-d7a9-427f-b76c-65eab58e8a06" />
Introduction
Godot AI Chat is an AI chat plugin deeply integrated into the Godot editor. Besides standard AI chat features, it also provides Agent capabilities implemented based on the ReAct framework.
Core Features
- Agent Experience: The
ReActagent can autonomously initiate multi-turn dialogue execution through tool calls, which is crucial for building automated workflows. - Smart Drag & Drop: The prompt input box accepts user prompts and supports dragging files/folders into it, automatically parsing them into project paths
res://xxxxx. - Smart Conversion: When the user's prompt contains image paths or scene file paths, the plugin automatically converts them internally into actual images and markdown-formatted scene tree structures.
- Multimodal Support: Supports sending images to models with vision capabilities (model support required).
- Context Sliding Window: When the number of session turns exceeds the set maximum, the plugin trims the session context based on the set limit before sending, ensuring the context content does not exceed the model's context window.
- Skill System: Based on Anthropic's
Skillconcept, this plugin implements its own on-demand loading skill system. It strictly limits mounting to only 1 skill pack at a time, maximizing the maintenance of the model's context window usage. - Security Mechanism: The tool base class
ai_tool.gdprovides path blacklist constants and path blacklist checking methods. This not only prevents AI from accidentally modifying critical files but also makes it convenient for developers to customize project paths they want to lock. - Compliant Write Operations: All tools with write functionality operate on files through engine APIs, ensuring the compliance of file operations.
- Transparent Invocation: The chat window supports real-time display of AI tool call details and explicitly shows the call parameters for each tool, making it convenient for developers to detect erroneous AI tool behavior in time to stop it.
- Real-time Feedback: All operations on scenes and script files are reflected in real-time on the editor's relevant interfaces, facilitating real-time supervision by developers.
- Extensibility: The plugin's architecture achieves maximum module decoupling, making the integration of new API providers and the development of new AI tools very friendly. Errors won't break the plugin's core functionality, which is crucial for real-time debugging of new modules.
A ReAct agent is an AI agent that uses the “reasoning and acting” (ReAct) framework to combine chain of thought (CoT) reasoning with external tool use. The ReAct framework enhances the ability of a large language model (LLM) to handle complex tasks and decision-making in agentic workflows.
Installation & Enablement
⚠️ Note: This plugin only supports Godot 4.5 and above.
1. Download Dependencies: - Download this plugin: Godot AI Chat - Download the dependency plugin: Context Toolkit > Context Toolkit is used to provide file context APIs; this plugin's default tool get_context depends on it.
2. Installation: - Unzip both plugin packages. - Drag the godot_ai_chat and context_toolkit folders into the project's addons/ directory.
3. Enablement: - Open the Godot editor, go to Project -> Project Settings -> Plugins. - Check to enable Godot AI Chat and Context Toolkit (Optional).
Configuration Guide
When enabled for the first time, a red error message Please Configure Plugin Settings will appear in the status bar of the chat interface. This indicates that relevant configurations need to be completed in the Settings panel of the plugin interface.
1. API Provider Settings
This plugin supports various API standards:
| API Type | Description | Representative Providers |
|---|---|---|
| **OpenAI Compatible** | The most universal standard interface | OpenAI Official, DeepSeek, Kimi (Moonshot), SiliconFlow, OpenRouter, LM Studio (Local), Ollama (Local) |
| **Gemini** | Google native interface | Google Gemini Official |
| **Anthropic Compatible** | Claude series interface | Anthropic Official, some proxy providers |
⚠️ Note: ZhipuAI in the API providers is actually a non-standard OpenAI compatible API, not a special API type.
⚠️ Note: Coding Plan subscription services provided by model vendors are usually only for their specified clients and most likely cannot be used with this plugin. Please be aware.
4. Other Key Settings
- Tavily API Key: The model can only use the
search_webweb search tool normally after configuring this. - Max Chat Turns: Default
12. Controls the number of dialogue turns sent to the model. Too high may consume a large number of Tokens or trigger limits; too low may lead to context loss. - Temperature: Default
0.8. Lower values result in more rigorous answers (suitable for coding), while higher values result in more divergent answers (suitable for creativity). - Log Level: Default only checks
WarnandError. Generally no need to change unless you want to view the plugin's debug status in the editor's Output in real-time. - System Prompt: The system prompt is the first special message sent to the AI model. Unlike normal conversation messages, the model gives higher weight and attention to the system prompt, ensuring all subsequent replies follow its settings.
Tips: Max Chat Turns can be adjusted dynamically as needed during the session.
Tips: The plugin repository includes a default system prompt document system_prompt.md that has undergone multiple iterations. It works out of the box and is recommended for users new to this plugin. You can then make personalized adjustments based on actual needs.
2. Base URL (API Address)
The interface address pointing to the AI service provider.
<details> <summary>Click to view common Base URL list</summary>
- Local Run (No Key required)
- LM Studio:
http://127.0.0.1:1234 - Ollama:
http://127.0.0.1:11434/v1 - Remote Services
- OpenRouter:
https://openrouter.ai/api - Google Gemini:
https://generativelanguage.googleapis.com - Moonshot (Kimi):
https://api.moonshot.cn - SiliconFlow:
https://api.siliconflow.cn - DeepSeek:
https://api.deepseek.com - ZhipuAI:
https://open.bigmodel.cn/api/paas/v4/
Note: If your provider is not in the list, please look for "API Endpoint" or "Base URL" in their documentation. </details>
3. API Key
- Remote API Service: Required.
- Local API Service (e.g., LM Studio): Usually can be left empty.
Chat Interface Overview
<img width="505" height="836" alt="截屏2026-03-01 23 35 31" src="https://github.com/user-attachments/assets/c97b0f37-98cc-4cef-bfe5-05d2d351b1c9" />
The interface contains from top to bottom: 1. Status Bar: Displays plugin working status, error messages, etc., in real-time. 2. Session Turns: Real-time monitoring of current session turns and the maximum set session turns. 3. Token Monitor: Displays current dialogue and cumulative historical Token consumption, helping you control costs. 4. Function Area: - Model Hot-Switch: Quickly switch between different models. - Session Management: Create, load, and delete sessions. 5. Prompt Input Box: Supports text input and file drag-and-drop referencing. 6. Export and Send: Save dialogue records as Markdown and the send button.
- 构建多智能体协作系统的 Agent 开发者
- 本地部署优先选 GGUF 量化模型，节省显存并保持响应速度
- Agent 任务先做 dry-run 验证工具调用链，再开启自主执行
- API key 直接提交到 git 仓库（请用 .env 并加入 .gitignore）
- 显存不足直接 OOM — 优先降低 context 或换更小的量化模型
- 本地部署：CPU 8GB 起，GPU 推荐 16GB+ 显存
- 云端托管：可放在 Vercel / Railway / Fly.io 等 PaaS 平台
⚡ 核心功能
- 开源免费，支持本地部署，数据完全自主可控
- 活跃的 GitHub 开源社区，持续迭代更新
- 提供详细文档和使用示例，新手友好
- 支持自定义配置，灵活适配不同使用环境
- 可作为基础组件集成进现有技术栈或进行二次开发
- 构建多智能体协作系统的 Agent 开发者
- 本地部署优先选 GGUF 量化模型，节省显存并保持响应速度
- Agent 任务先做 dry-run 验证工具调用链，再开启自主执行
- API key 直接提交到 git 仓库（请用 .env 并加入 .gitignore）
- 显存不足直接 OOM — 优先降低 context 或换更小的量化模型
👥 适合人群
🎯 使用场景
- 本地部署运行，保护数据隐私，满足合规要求
- 自定义集成到现有系统，扩展技术栈能力
- 作为开源基础组件进行商业化二次开发
⚖️ 优点与不足
- +MIT 协议，可免费商用
- +完全开源免费，无授权费用
- +本地部署，数据完全自主可控
- +开发者社区支持，遇问题可查可问
- −安装和初始配置可能需要一定技术基础
- −功能完整性通常不如成熟商业产品
- −技术支持主要依赖开源社区，响应速度不稳定
AI Skill Hub 为第三方内容聚合平台，本页面信息基于公开数据整理，不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后，再部署至生产环境，并做好必要的安全评估。
✅ MIT 协议 — 最宽松的开源协议之一，可自由商用、修改、分发，仅需保留版权声明。
🔗 相关工具推荐
❓ 常见问题 FAQ
总体来看，Godot AI聊天 是一款质量良好的AI工具，在同类工具中具备一定竞争力。AI Skill Hub 将持续追踪其更新动态，建议收藏备用，结合自身场景选择合适时机引入使用。
| 原始名称 | Godot-AI-Chat |
| 原始描述 | 开源AI工具：Godot AI Chat is a Godot plugin that enables direct conversation with LLMs withi。⭐18 · GDScript |
| Topics | GodotAILLM聊天 |
| GitHub | https://github.com/snougo/Godot-AI-Chat |
| License | MIT |
| 语言 | GDScript |
收录时间：2026-05-29 · 更新时间：2026-05-30 · License：MIT · AI Skill Hub 不对第三方内容的准确性作法律背书。