Mapbox Web Performance Patterns

Mapbox Performance Patterns Skill

This skill provides performance optimization guidance for building fast, efficient Mapbox applications. Patterns are prioritized by impact on user experience, starting with the most critical improvements.

Performance philosophy: These aren’t micro-optimizations. They show up as waiting time, jank, and repeat costs that hit every user session.

Priority Levels

Performance issues are prioritized by their impact on user experience:

• 🔴 Critical (Fix First): Directly causes slow initial load or visible jank
• 🟡 High Impact: Noticeable delays or increased resource usage
• 🟢 Optimization: Incremental improvements for polish

🔴 Critical: Eliminate Initialization Waterfalls

Problem: Sequential loading creates cascading delays where each resource waits for the previous one.

Note: Modern bundlers (Vite, Webpack, etc.) and ESM dynamic imports automatically handle code splitting and library loading. The primary waterfall to eliminate is data loading - fetching map data sequentially instead of in parallel with map initialization.

Anti-Pattern: Sequential Data Loading

// ❌ BAD: Data loads AFTER map initializes
async function initMap() {
  const map = new mapboxgl.Map({
    container: 'map',
    accessToken: MAPBOX_TOKEN,
    style: 'mapbox://styles/mapbox/streets-v12'
  });

  // Wait for map to load, THEN fetch data
  map.on('load', async () =>

Timeline: Map init (0.5s) → Data fetch (1s) = 1.5s total

Solution: Parallel Data Loading

// ✅ GOOD: Data fetch starts immediately
async function initMap() {
  // Start data fetch immediately (don't wait for map)
  const dataPromise = fetch('/api/data').then((r) => r.json());

  const map = new mapboxgl.Map({
    container: 'map'

Timeline: Max(map init, data fetch) = ~1s total

Set Precise Initial Viewport

// ✅ Set exact center/zoom so the map fetches the right tiles immediately
const map = new mapboxgl.Map({
  container: 'map',
  style: 'mapbox://styles/mapbox/streets-v12',
  center: [-122.4194, 37.7749],
  zoom: 13
});

// Use 'idle' to know when the initial viewport is fully rendered
// (all tiles, sprites, and other resources are loaded; no transitions in progress)
map.once

If you know the exact area users will see first, setting center and zoom upfront avoids the map starting at a default view and then panning/zooming to the target, which wastes tile fetches.

Defer Non-Critical Features

// ✅ Load critical features first, defer others
const map = new mapboxgl.Map({
  /* config */
});

map.on('load', () => {
  // 1. Add critical layers immediately
  addCriticalLayers(map);

  // 2. Defer secondary features
  // Note: Standard style 3D buildings can be toggled via config:
  // map.setConfigProperty('basemap', 'show3dObjects', false);
  requestIdleCallback

Impact: Significant reduction in time-to-interactive, especially when deferring terrain and 3D layers

🔴 Critical: Optimize Initial Bundle Size

Problem: Large bundles delay time-to-interactive on slow networks.

Note: Modern bundlers (Vite, Webpack, etc.) automatically handle code splitting for framework-based applications. The guidance below is most relevant for optimizing what gets bundled and when.

Style JSON Bundle Impact

// ❌ BAD: Inline massive style JSON (can be 500+ KB)
const style = {
  version: 8,
  sources: {
    /* 100s of lines */
  },
  layers: [
    /* 100s of layers */
  ]
};

// ✅ GOOD: Reference Mapbox-hosted styles
const map = new mapboxgl.Map({
  style: 'mapbox://styles/mapbox/streets-v12' // Fetched on demand

Impact: Reduces initial bundle by 30-50% when moving from inlined to hosted styles

🟡 High Impact: Optimize Marker Count

Problem: Too many markers causes slow rendering and interaction lag.

Performance Thresholds

• < 100 markers: HTML markers OK (Marker class)
• 100-10,000 markers: Use symbol layers (GPU-accelerated)
• 10,000+ markers: Clustering recommended
• 100,000+ markers: Vector tiles with server-side clustering

Anti-Pattern: Thousands of HTML Markers

// ❌ BAD: 5,000 HTML markers = 5+ second render, janky pan/zoom
restaurants.forEach((restaurant) => {
  const marker = new mapboxgl.Marker()
    .setLngLat([restaurant.lng, restaurant.lat])
    .setPopup(new mapboxgl.Popup().setHTML(restaurant.name))
    .addTo(map);
});

Result: 5,000 DOM elements, slow interactions, high memory

Solution: Use Symbol Layers (GeoJSON)

// ✅ GOOD: GPU-accelerated rendering, smooth at 10,000+ features
map.addSource('restaurants', {
  type: 'geojson',
  data: {
    type: 'FeatureCollection',
    features: restaurants.map((r) => ({
      type: 'Feature',
      geometry: { type: 'Point'

Performance: 10,000 features render in <100ms

Solution: Clustering for High Density

// ✅ GOOD: 50,000 markers → ~500 clusters at low zoom
map.addSource('restaurants', {
  type: 'geojson',
  data: restaurantsGeoJSON,
  cluster: true,
  clusterMaxZoom: 14, // Stop clustering at zoom 15
  clusterRadius: 50 // Radius relative to tile dimensions (512 = full tile width)
});

Impact: 50,000 markers at 60 FPS with smooth interaction

Summary: Performance Checklist

When building a Mapbox application, verify these optimizations in order:

🔴 Critical (Do First)

• Not done: Load map library and data in parallel (eliminate waterfalls)
• Not done: Use dynamic imports for map code (reduce initial bundle)
• Not done: Defer non-critical features (terrain, custom 3D layers, analytics)
• Not done: Use symbol layers for > 100 markers (not HTML markers)
• Not done: Implement viewport-based data loading for large datasets

🟡 High Impact

• Not done: Debounce/throttle map event handlers (geocode inputs, moveend)
• Not done: Optimize queryRenderedFeatures with layers filter and bounding box
• Not done: Use GeoJSON for < 5 MB, vector tiles for > 20 MB
• Not done: Always call map.remove() on cleanup in SPAs / page teardown
• Not done: Attach map.on('error', …) (or visible error UI) so style/tile/token failures are not silent
• Not done: Reuse popup instances (don’t create on every interaction)
• Not done: Use feature state instead of dynamic layers for hover/selection
• Not done: Cluster demos: generate enough points to stress clustering (thousands, not a few hundred)

Agent anti-pattern: happy-path only

First-pass agent code often ships a map with no map.on('error'), no map.remove(), and a tiny point set that never exercises cluster: true. Production demos need error visibility, teardown, and realistic scale.

🟢 Optimization

• Not done: Consolidate multiple layers with data-driven styling
• Not done: Add mobile-specific optimizations (circle layers, disabled rotation)
• Not done: Set minzoom/maxzoom on layers to avoid rendering at irrelevant zoom levels
• Not done: Avoid enabling preserveDrawingBuffer or antialias unless needed

Measurement

// Measure initial load time
console.time('map-load');
map.on('load', () => {
  console.timeEnd('map-load');
  // isStyleLoaded() returns true when style, sources, tiles, sprites, and models are all loaded
  console.log('Style loaded:', map.isStyleLoaded());
});

// Monitor frame rate
let

Target metrics:

• Time to Interactive: < 2 seconds on 3G
• Frame Rate: 60 FPS during pan/zoom
• Memory Growth: < 10 MB per hour of usage
• Bundle Size: < 500 KB initial (map lazy-loaded)

Reference Files

For detailed patterns on specific topics, load the corresponding reference file:

• references/data-loading.md — GeoJSON vs Vector Tiles decision matrix, viewport-based loading, progressive loading, vector tiles for large datasets
• references/interactions.md — Debounce/throttle events, optimize feature queries, batch DOM updates
• references/memory.md — Map cleanup patterns, popup/marker reuse, feature state vs dynamic layers
• references/mobile.md — Device detection, mobile-optimized layers, touch interaction, constructor options
• references/layers-styles.md — Consolidate layers with data-driven styling, simplify expressions, zoom-based visibility