mold-ai 0.2.0

Local AI image generation CLI — FLUX, SDXL, SD3.5, Z-Image diffusion models on your GPU
use anyhow::Result;
use mold_core::Config;
use std::path::PathBuf;

use crate::theme;

pub async fn run(port: u16, bind: &str, models_dir: Option<String>) -> Result<()> {
    let config = Config::load_or_default();

    let models_path = match models_dir {
        Some(dir) => PathBuf::from(dir),
        None => config.resolved_models_dir(),
    };

    // Ensure models directory exists
    std::fs::create_dir_all(&models_path)?;

    println!(
        "{} Starting mold server on {}:{}",
        theme::icon_ok(),
        bind,
        port,
    );
    println!(
        "{} Models directory: {}",
        theme::icon_ok(),
        models_path.display(),
    );

    mold_server::run_server(bind, port, models_path).await
}