Performance and blocking

Overview

Tuning is iterative: start with defaults, measure X-ScreenNabster-Duration from sync responses, then introduce wait_until, wait, and blocking options until variance drops. Always validate visually—blocking scripts can break SPAs entirely.

Cache repeated captures

Set cache=true (or cached=true) when repeated requests should reuse a stored render. Cache hits return X-ScreenNabster-Cache: HIT, include X-ScreenNabster-Cache-Url when available, and do not consume monthly quota. Use cache_key to separate variants and cache_ttl to control signed URL lifetime. Fresh JSON responses include cache_url after a successful cache write; JSON cache hits return the cached image bytes with empty metadata because page metadata is not stored in the cache object.

curl -G "https://screennabster.com/api/v1/capture" \
  --data-urlencode "url=https://example.com" \
  --data-urlencode "cache=true" \
  --data-urlencode "cache_ttl=14400" \
  -H "X-API-Key: pk_live_…" \
  --output cached.png

Waits & readiness

wait_until picks the navigation milestone. networkidle is stricter than load and helps SPAs settle. Add wait (milliseconds) for fixed delays such as carousel animations, or wait_for_selector when a known selector marks readiness.

{
  "url": "https://example.com/app",
  "wait_until": "networkidle",
  "wait_for_selector": "[data-ready=true]",
  "wait": 250,
  "output": "png"
}

Block resource types

block_resources accepts Playwright types: image, stylesheet, font, media, script, xhr, fetch. Blocking images speeds up text-heavy captures; blocking scripts may blank React apps—test carefully.

curl -G "https://screennabster.com/api/v1/capture" \
  --data-urlencode "url=https://example.com" \
  --data-urlencode "block_resources=image,media" \
  -H "X-API-Key: pk_live_…" \
  --output lite.png

Content filters

filters enables higher-level bundles: cookie_banners, ads, popups, chats, trackers. They complement—not replace—your own hide_selectors lists for brand-specific overlays.

Strip ads

strip_ads: true flips on host-based blocking tuned for common ad networks. Pair with filters when pages still leak sponsored modules through first-party domains.

Tradeoffs

• Aggressive blocking can break A/B tests that rely on client-side scripts.
• For captures that still exceed your HTTP timeout, switch to background mode.
• Keep an eye on quota—every attempt counts once admitted past the gate.