Problem Livepeer 2.0 has active Agent and network-readiness work, but I could not find public results for a repeatable external paid-demand pilot and recurring fees reaching independent operators. The Network Engineering SPE mentions OpenRouter integration as an example of measurable impact; a contributor reports first paid outside-app calls through the Modules Gateway. What is the path from these early signals to durable commercial volume?
Why it matters GPU capacity only becomes a sustainable service when buyers can reliably purchase a defined capability and operators are paid from recurring usage. Standard API workloads and long-running live video will likely need different sales channels.
Proposed pilot and success criteria
Route one aggregator-compatible capability, such as an OpenAI-compatible vision-language endpoint, through OpenRouter or a similar paid distributor.
Separately validate an SLA-backed enterprise channel for real-time video, including reserved capacity and support.
Report monthly external paying customers, repeat usage, paid request volume, completion rate, p95 latency, gross paid revenue, and net fees paid to independent operators. Break out internal/test traffic, grants, centrally sourced APIs, and pass-through.
Open questions What blocks the pilot today: API format, metering and invoicing, privacy terms, QoS/capacity, or demand? Who owns each channel, and what are its go/no-go gates?
