sveltekit-modal

sveltekit-modal

Write Python endpoints in SvelteKit using Modal.

Why

• Start deploying Python endpoints in just a few steps
• GPU support for an entire ML Stack in your SvelteKit app
• Integrates completely, write +server.py just like your +server.js files
• Deploy the rest of your app anywhere, with SvelteKit's adapters for Vercel, Netlify, Cloudflare, etc.
• Severless Python with ease

Add to your SvelteKit project

• Install npm i -D sveltekit-modal
• Sign up for modal.com, the serverless python platform. All users get $30 free monthly credits!
• Update vite.config.js

  import { sveltekit } from '@sveltejs/kit/vite';
  import { defineConfig } from 'vite';
  
  import { sveltekit_modal } from 'sveltekit-modal/vite'; //Add this import
  
  export default defineConfig({
      plugins: [sveltekit_modal(), sveltekit()] //Add the `sveltekit_modal()` plugin
  });
  

• Update svelte.config.js

  import adapter from '@sveltejs/adapter-auto';
  import { vitePreprocess } from '@sveltejs/kit/vite';
  
  /** @type {import('@sveltejs/kit').Config} */
  const config = {
      preprocess: vitePreprocess(),
      kit: {
    	  adapter: adapter(),
    	  moduleExtensions: [".js", ".ts", ".py"] //Add this line, to resolve +server.py endpoints
      }
  };
  
  export default config;
  

• Create sveltekit_modal_config.py. The option stub_asgi is passed to Modal. This is where you can define GPU acceleration, secrets, and an Image for pip installs, etc. Explore their docs!

  import modal
  import sveltekit_modal
  
  config: sveltekit_modal.Config = {
    'log': False,
    'stub_asgi': {
        'image': modal.Image.debian_slim()
    }
  }
  

• Update .gitignore, add !.env.production.
• Write your endpoints! See an example here.

Use

• Develop like a normal SvelteKit app, just npm run dev
• Deploy all your python endpoints with one command, just npx sveltekit-modal deploy