FastAgents compute dashboard

Real compute.
In full view.

Video-AI activity and hardware, at a glance. Reported metrics are not a guarantee of current availability.

Reported AI lanes --

Advertised video-AI lanes, not a physical GPU count.

Streams served, 24h --

Reported video streams over the last day.

Estimated visitors, 30 days --

Includes the homepage. Repeat visits and bots may affect the estimate.

Power consumption, 24h --

Estimated fleet GPU energy use from the 24h power curve.

Daydream video-to-video

Video-AI activity

Reported workloads. Availability varies.

Output FPS, 24h --

Average of recent output samples.

Hourly jobs, 3h avg --

Reported hourly jobs, smoothed over three hours.

Public endpoints

Next on our roadmap

Livepeer Agents Upcoming

Agent workloads are on our roadmap. Details to follow.

Input tokens, 24h Upcoming

Agent usage reporting is upcoming. Not live telemetry.

Output tokens, 24h Upcoming

Usage reporting is coming after launch.

Peak tok/sec, 24h Upcoming

No agent throughput is claimed before launch.

Agent jobs, 24h Upcoming

Agent job reporting is not enabled.

Agent models Upcoming

Model lineup to be announced.

GPU inventory

GPU hardware

Reported hardware, grouped by GPU family and memory.

Overall GPU power consumption --

Reported GPU power draw.

Website activity

Visitors around the world

Last 30 days

Page visits--
Countries shown--
Map detailCountry

Loading visitor totals...