Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.livepeer.org/llms.txt

Use this file to discover all available pages before exploring further.


Livepeer pricing is decentralised: orchestrators advertise prices per pipeline type and model, and gateways filter by their configured maximum. There is no central price list. The prices below reflect community gateway defaults and typical orchestrator ranges as of May 2026.

Billing Units

All prices are denominated in wei (ETH on Arbitrum One). The TicketBroker contract settles winning tickets; the expected payment per ticket equals faceVal × winProb.

Community Gateway

The community gateway at dream-gateway.livepeer.cloud is operated by the Livepeer Foundation for development access. It routes to whichever orchestrators in the active set are warm for the requested pipeline. The community gateway does not publish a fixed price schedule. It accepts jobs within the orchestrator price range active on the network at request time. For development and low-volume testing this is adequate; for production workloads with cost predictability requirements, operate a self-hosted gateway with -maxPricePerUnit configured. Rate limiting: The community gateway applies rate limiting to unauthenticated requests. Sustained high-volume traffic will be throttled. There is no published rate limit figure; if requests begin returning 429 responses, reduce request frequency or move to a dedicated gateway.

Discovering Current Prices

Orchestrators advertise AI pipeline pricing in aiModels.json and register this data via the AI Service Registry on Arbitrum One. Gateways query this at session establishment. To find current prices:
  1. Network capability dashboard: tools.livepeer.cloud/ai/network-capabilities shows warm pipelines and model IDs across the active orchestrator set, but does not display per-orchestrator pricing
  2. GetOrchestratorInfo RPC: gateways call this to retrieve price and capability data directly from orchestrators during session setup
  3. Livepeer Explorer: explorer.livepeer.org shows historical per-unit prices for transcoding orchestrators

Orchestrator Pricing Reference

Orchestrators configure pricing per pipeline in aiModels.json. The price_per_unit field is in wei and its meaning varies by pipeline: For transcoding, -pricePerUnit is set in wei per pixel at node startup, not in aiModels.json.
Setting price_per_unit: 0 makes the orchestrator accessible at any gateway’s configured maximum price. Use 0 for development and test environments; set a positive value in production to recover infrastructure costs.

Probabilistic Micropayments

How prices translate to PM ticket face values and win probabilities.

Per-Second Compute

Per-second billing model for the live-video-to-video pipeline.

Orchestrator Selection

How gateways filter orchestrators by price and capability.

ETH Escrow

TicketBroker deposits that cover expected payment for jobs submitted.
Last modified on May 19, 2026