Live benchmarks are available at
benchmark.scrape.st — updated in real-time.
At a Glance
What We Measure
Latency
Time from when a post is created on the source platform to when Scrapest delivers it to your webhook/stream. Measured end-to-end across all 3 web-push shards.- Source Latency — Platform → Scrapest ingestion
- Internal Latency — Scrapest processing → webhook/stream delivery
- P50 / P95 / P99 — Percentile distribution over a 24-hour rolling window
Coverage
Scrapest tracks 3,000+ accounts simultaneously across Twitter/X with zero rate-limit throttling, using a distributed account pool with geographic proxy distribution (GB/US/KW).Latency Comparison
Scrapest
X API (Official)
The official X API v2 search endpoint has index lag of 5-15 seconds. Filtered stream (Enterprise only at $42,000+/year) claims ~seconds latency but is inaccessible to most developers.Cost Comparison
For a typical integration tracking 100 accounts with real-time delivery:Live Metrics
Query real-time latency data directly from the API:Metrics endpoints are public — no authentication required. All data uses a
24-hour rolling window.
Why It Matters
- Trading bots need sub-second latency — Scrapest delivers in ~200ms vs X API’s 5-15s
- Social analytics need full coverage — Scrapest has zero rate limits vs X’s 100-500 req/15min caps
- Multi-platform monitoring — one API for Twitter, Telegram, and Discord instead of 3 separate integrations
Get started: Core · REST API · Streams