This dashboard provides a comprehensive view of Firecrawl SDK usage using trace data. It is built on the firecrawl.* span attributes exported by the firecrawl-otel instrumentation and focuses on call volume, error rate, latency, credits consumed, and results returned, all broken down per operation (scrape, search, map, crawl, and batch_scrape). Every panel filters on firecrawl.operation EXISTS rather than a fixed service name, so the dashboard renders for both short-lived jobs and long-running services.
Dashboard Preview
Dashboards → + New dashboard → Import JSON
What This Dashboard Monitors
This dashboard tracks usage, performance, and cost metrics for your Firecrawl workloads using OpenTelemetry trace data to help you:
- Track Credit Consumption: Sum
firecrawl.credits_usedoverall and per operation to see where spend is concentrated against your Firecrawl plan. - Monitor Scraping Volume: Watch call counts over time and per operation to understand traffic patterns and which operations dominate.
- Catch Errors Early: Surface the share of calls that fail as a single percentage for an at-a-glance health signal.
- Monitor Latency: Watch p50, p95, and p99 duration to surface both typical performance and the tail latency that shapes worst-case experience.
- Compare Operations: Identify which operations are slowest under load and how they differ in cost and yield.
- Measure Yield: Track how many results each operation returns to spot drops that do not show up as outright errors.
- Inspect Recent Calls: Drill into the most recent Firecrawl spans when investigating an error or a latency spike.
Panels Included
Usage Summary (Top Row)
| Panel | Type | What It Shows |
|---|---|---|
| Total Firecrawl Calls | Value | Count of all Firecrawl operation spans in the selected window, a quick pulse on how much scraping work is running through the SDK |
| Error Rate | Value | Share of calls that failed, computed as errored spans divided by all spans and rendered as a percentage |
| Total Credits Used | Value | Sum of firecrawl.credits_used across all calls, tracking consumption against your Firecrawl plan |
| Total Results Returned | Value | Sum of firecrawl.results across all calls, showing how much content the operations bring back |
Call Volume
- Calls Over Time by Operation: Time-series graph of call counts grouped by
firecrawl.operation, revealing traffic patterns and which operations dominate. - Call Distribution by Operation: Pie chart of each operation's share of total calls, complementing the trend graphs with a clear share-of-traffic view.
Latency
- Latency Percentiles (p50 / p95 / p99): Time-series graph of duration percentiles across all Firecrawl spans, surfacing both typical performance and tail latency.
- p95 Latency by Operation: Time-series graph of 95th-percentile duration per operation, identifying which operations are slowest under load.
Credits & Results
- Credits Used Over Time by Operation: Time-series graph of credits consumed grouped by operation, showing where spend is concentrated and how it trends.
- Results Returned Over Time by Operation: Time-series graph of results returned grouped by operation, useful for spotting drops in yield that do not show up as outright errors.
Per-Operation Breakdown
- Operation Summary: Table of call count, p95 latency, average latency, results, and credits per operation, giving a full side-by-side comparison across
scrape,search,map,crawl, andbatch_scrape.
Recent Activity
- Recent Firecrawl Calls: List of the most recent Firecrawl spans ordered by timestamp descending, for drilling into individual calls when investigating an error or a latency spike.
