live · public · aggregate only

Network statistics

How the public ChainLayer2 gateway is doing right now: uptime from our own minute-by-minute probes, request volume and latency, the models people use, and the state of each service.

Loading… refreshes every 60s

Version

—

current release

Uptime, 30 days

—

gateway probes

Requests, 24h

—

all models

Tokens, 24h

—

in + out, rounded

Avg latency, 24h

—

end to end

Models online

—

in the catalog

Providers healthy

—

circuit closed

Traffic

hourly buckets, your local time

Requests

per hour

—

View as table

Median latency (p50)

successful requests, end to end

—

View as table

Top models

share of requests

Loading…

Services

checked every minute from the server

Model gateway/v1

90 days agotoday
Remote relay/relay

90 days agotoday
Accountssign-in

90 days agotoday

API

/stats/v1/Index: every endpoint and the schema version.
/stats/v1/summaryThe headline numbers on this page.
/stats/v1/timeseries?range=24h|7d|30dBucketed requests, tokens, errors, p50 latency.
/stats/v1/models?range=24h|7d|30dModel share of requests and p50 latency.
/stats/v1/healthEach service up or down, with its last checks.
/stats/v1/uptimeRolling 30- and 90-day uptime, plus per-day uptime.

Aggregate only. These numbers never include API keys or key names, users or emails, IP addresses, prompts or responses, or provider accounts. Models with fewer than five requests in the window are grouped into “other”, and token counts are rounded. JSON, CORS-open, cached for 60 seconds.