Skip to content
Last Updated
03:46 3 thg 11, 2026

Connecting Local AI Models (Ollama, LM Studio) ​

Mục này chưa có bản dịch Tiếng Việt

Nội dung đang được hiển thị bằng tiếng Anh. Nếu bạn muốn đóng góp bản dịch, vui lòng tham gia tại kho lưu trữ xmcl-page trên GitHub.

XMCL lets you connect locally running LLMs executing directly on your own computer's GPU or CPU. This provides:

  • Absolute Privacy: Zero telemetry, prompts, or Minecraft crash dumps leave your local network.
  • 100% Free: No recurring subscriptions, token metering, or quota exhaustion.
  • Offline Capability: The assistant functions even when your internet connection is down.

Option 1: Ollama (Recommended & Simplest)

Ollama is the leading open-source tool for running local AI models across Windows, macOS, and Linux.

Step 1. Install Ollama and pull a model ​

  1. Download and run the installer from ollama.com.
  2. Open a terminal and download your desired model:
    bash
    # Lightweight model for standard laptops (~4 GB RAM):
    ollama run llama3.2
    
    # Excellent model for code, mods, and reasoning (~8 GB VRAM / RAM):
    ollama run qwen2.5:7b
  3. Ensure the Ollama background daemon is running (indicated by the llama icon in your system tray).

Step 2. Configure XMCL ​

  1. Open XMCL -> Settings -> General.
  2. Toggle on Developer Mode.
  3. In the AI Agent section, enter:
    • Endpoint: http://localhost:11434/v1/chat/completions
    • Model: llama3.2 (or qwen2.5:7b, matching your pulled Ollama model tag).
    • API Key: Enter any placeholder string like ollama (Ollama ignores authorization, but XMCL expects a non-empty field).

Option 2: LM Studio

LM Studio is an intuitive desktop app for discovering, downloading, and hosting quantized GGUF models locally.

Step 1. Start the local server in LM Studio ​

  1. Download the app from lmstudio.ai.
  2. Search and download a model (e.g. Qwen2.5-7B-Instruct or Llama-3.2-3B-Instruct).
  3. Click the Developer / Local Server tab on the left sidebar.
  4. Select your downloaded model from the top dropdown.
  5. Click Start Server (defaults to port 1234).

Step 2. Configure XMCL ​

  1. In XMCL Settings -> General -> AI Agent, configure:
    • Endpoint: http://localhost:1234/v1/chat/completions
    • Model: Copy the model identifier displayed in LM Studio.
    • API Key: lm-studio (or any string).

Hardware Sizing & Recommended Models

Local inference speed depends heavily on your GPU VRAM or system RAM:

Model SizeRecommended HardwareBest Models for Minecraft Tasks
1B – 3B ParametersAny modern laptop, 6–8 GB RAMllama3.2:1b, llama3.2:3b, qwen2.5:3b
7B – 8B ParametersGPU with 6–8 GB VRAM (RTX 3060/4060) or 16 GB RAMqwen2.5:7b, llama3.1:8b, mistral:7b
14B+ ParametersDedicated GPU with 12–16+ GB VRAMqwen2.5:14b, deepseek-coder-v2

Best Model for Crash Logs

The qwen2.5 model series (especially qwen2.5:7b) exhibits the highest accuracy when parsing Java crash stack traces, identifying missing dependencies, and suggesting mod resolutions among compact open-weights models.


Troubleshooting Local Setups

  • Connection Refused (ECONNREFUSED): The local inference daemon is not active. Verify that Ollama or LM Studio's server is running in the background before querying.
  • Slow Responses / High Latency: The model is too large for your GPU memory and is spilling layers onto the CPU. Switch to a smaller parameter model (like llama3.2:3b) or a 4-bit quantized version (q4_k_m).
  • Model not found: Double check the model tag spelling in XMCL settings against your running server.