Local AI Generation MCP Server (local_ai_gen)
This project runs a local MCP (Model Context Protocol) server that exposes tools for:
- Text to Image
- Text to Music / Audio
- Text to Speech
- Image/Text to 3D Model
Power your agent
การติดตั้งจากผู้เผยแพร่ ไม่ใช่เอนด์พอยต์ของ PlugMCP
URL แบบรีโมตหรือคำสั่งบนเครื่องของคุณ? การติดตั้งสองแบบต่างกันอย่างไร
รันบน PlugMCP
A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.
เปิดจากผู้เผยแพร่
local_ai_gen)This project runs a local MCP (Model Context Protocol) server that exposes tools for:
Local AI asset generation MCP server: end-to-end text-to-image/audio/speech and image/text-to-3D with Qwen3-TTS, Stable Audio Open, SSD-1B, and TripoSR.
segmind/SSD-1B
A fast, distilled version of SDXL that works well on consumer GPUs.stabilityai/stable-audio-open-1.0
A high-quality open model for generating sound effects, short music tracks, and ambient audio. (Note: Requires accepting license terms)Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice
A robust, multilingual text-to-speech model.stabilityai/TripoSR
A fast feed-forward model for image-to-3D reconstruction.stable-audio-open-1.0)The Stable Audio Open model is gated and requires you to accept its license before downloading.
pip install -U "huggingface_hub"
hf auth
(Paste your token when prompted. You do not need to add it as a git credential).python -m venv .venv
source .venv/bin/activate
python setup.py
What python setup.py does:
generated_images, generated_audio, generated_models, models)third_party/TripoSR if it is missingrequirements.txttorchmcubes against your active torch version (needed for 3D generation)flash-attn.To run the MCP server manually (via standard I/O):
python mcp_server/main.py
Note: The first run of any specific tool will be slower because it has to download the weights for that model into your Hugging Face cache.
You can run the included smoke test script to verify all models are working correctly:
python smoke_test_generate.py
generate_imagegenerate_audiogenerate_speechgenerate_3d_modelhealth_checkgenerated_images, generated_audio, and generated_models by default.generate_audio and generate_speech, you can override the destination with:output_dir (highest priority), orGENAI_OUTPUT_AUDIO_DIRgenerate_image, override destination with tool arg output_dir or env var GENAI_OUTPUT_IMAGE_DIRgenerate_3d_model, override destination with tool arg output_dir or env var GENAI_OUTPUT_MODEL_DIRTo use this server in an MCP-compatible client, add the following to your mcp.json (or the respective MCP configuration file for your client). Make sure to replace <YOUR_PROJECT_PATH> with the absolute path to where you cloned this repository:
{
"mcpServers": {
"local_ai_gen": {
"command": "<YOUR_PROJECT_PATH>/.venv/bin/python",
"args": [
"<YOUR_PROJECT_PATH>/mcp_server/main.py"
],
"env": {}
}
}
}
Local AI asset generation MCP server: end-to-end text-to-image/audio/speech and image/text-to-3D with Qwen3-TTS, Stable Audio Open, SSD-1B, and TripoSR.
เครื่องมือ
ยังไม่มีรายการเครื่องมือ เซิร์ฟเวอร์แบบรีโมตจะได้รายการหลังการตรวจสอบสถานะครั้งแรก
FAQ
คัดลอกคอนฟิก stdio สำหรับ Cursor, Claude Desktop, Claude Code หรือ VS Code ได้ในหน้านี้ คอนฟิกนี้ใช้การติดตั้งของผู้เผยแพร่ ไม่ใช่พร็อกซีของ PlugMCP
ระเบียนนี้ไม่ได้ระบุตัวแปรสภาพแวดล้อมที่จำเป็น
Local Asset Gen ทำงานแบบโลคัลผ่าน stdio
ตรวจสอบล่าสุดเมื่อ 2026-10-28 Local AI asset generation MCP server: end-to-end text-to-image/audio/speech and image/text-to-3D with Qwen3-TTS, Stable Audio Open, SSD-1B, and TripoSR.