Node Provider Remuneration

Node providers are paid for operating node machines on the Internet Computer. Reward rates are set by the Network Nervous System (NNS) and can only be changed by NNS proposal adopted by the community. This entry summarises the current rates, the formula behind the Gen-2 model, and the variants that apply to earlier hardware.

Minting cadence and currency

The NNS mints node provider rewards every 2,629,700 seconds (one-twelfth of a year). The current and next minting dates are published at next-rewards-gnj.caffeine.xyz↗.

Rewards are denominated in XDR (Special Drawing Rights, the international reserve asset issued by the IMF) and converted to ICP at the 30-day moving average ICP/XDR price at minting time. The conversion rate is tracked on the Internet Computer dashboard↗ under the Network Overview ’s “Conversion Rate” chart. Converted ICP is transferred to the principal that the provider registered during onboarding.

What rewards depend on

Three factors set the rate paid for any given node:

• Hardware generation — gen-1 or gen-2.
• Geographic location — capital and operating costs differ; underserved regions are weighted more.
• The provider’s total node count — later nodes are rewarded less, to discourage concentration of ownership.

A future model is expected to combine fixed rewards with automated incentives and penalties tied to node performance.

Initial Gen-1 model — type-0

Nodes purchased before genesis are gen-1, and all of them began on type-0:

Type-1 — the storage-upgrade rate

Soon after genesis the storage capacity of all gen-1 nodes was increased to support larger subnet states. Type-1 rates fold in the extra hardware cost and the country-by-country shipping and installation costs that providers absorbed.

Gen-1.1 — after the original 48 months

Gen-1 servers continuing to run beyond their initial 48-month agreements are paid under the Gen-1.1 schedule, adopted via NNS motion proposal #132553↗ (forum discussion here↗). Rewards are fixed and lower than type-1; a maximum of 42 nodes per provider applies; the model is expected to remain in force for roughly 24 months.

A subsequent motion proposal #135253↗ (forum thread↗) added a 10% bonus on top of the highest Gen-1.1 rate for providers who relocate their nodes to a non-EU member state.

Under the same proposal:

• Excess nodes may be sold to another provider while staying in the same data center.
• Excess nodes may be sold to an optimal provider and moved to a new country and data center.
• No provider may exceed 42 nodes.

The receiving-provider procedure is documented under Steps for Gen-1 onboarding after 48 months.

Gen-2 model

The gen-2 schedule is built on two principles:

• Higher rewards for the first nodes of a new provider, to attract ownership decentralization.
• Higher rewards in underserved geographies — South America, Africa, parts of Asia, Australia — to drive geographic decentralization.

The model is parameterised by:

• A geography multiplier mult(g). 2 for established regions (Europe, North America); 3 for regions with limited current presence (Africa, South America).
• A reduction coefficient r(np, g). The n-th node of a provider is rewarded with the base amount multiplied by r ^ (n-1). The coefficient is geography-dependent.

For a geography g, let cost(g) be the four-year total of capital and operating expense, in XDR. The monthly reward for the n-th node of provider np in geography g is:

reward(g, n) = cost(g) * mult(g) * r(np, g) ^ (n-1) / (4 * 12)

If a provider has nodes of more than one type or geography within a single country, the NNS sums the country’s total node count and applies the per-type rate to each.

The reduction coefficient r(np, g) is applied per (provider, country) pair: two providers in the same country each start at the 1st-node multiplier of 1 and are reduced independently as they add nodes.

Version 2 base rates

Worked example — first ten nodes in South Africa

reward(south africa, n)
  = cost(south africa) * mult(south africa) * r(south africa) ^ (n-1) / (4 * 12)
  = (21,455 + 22,531) * 3 * 0.98 ^ (n-1) / (4 * 12)
  = 2748 * 0.98 ^ (n-1)

In the original gen-2 analysis, additional nodes only generate positive incremental cash flow up to a region-dependent ceiling — roughly 15 nodes in Europe, only 2 in the US. New providers should run their own internal-rate-of-return and payback calculations against their own capex and opex; the published rates are based on estimates and real costs vary with vendor and data-center contracts.

Version 2.1 — current model from May 2023

Adopted after the discussion in this forum thread↗, Version 2.1 supersedes Version 2 with these changes:

• The generic Asia entry is removed; specific countries are listed individually.
• Hong Kong and India are added.
• The multiplier is set to 2 for all countries. A future proposal may differentiate countries by risk premium.
• A per-country cap of 50 nodes applies to new-country entries. Once reached, the reduction coefficient is tightened so only one or two further nodes can be added — the same dynamic as already applies in the US and Switzerland.

[!NOTE] As of the V2.1 update, network capacity is sufficient for the current canister load and for decentralization measured by the Nakamoto coefficient; no new country entries are being added. Any further changes will follow the IC Topology Roadmap discussion on the developer forum and a subsequent NNS proposal.

Reward calculator

A community-maintained calculator is available at calconic.com↗.

Trustworthy Node Metrics

Trustworthy Node Metrics were introduced in February 2024 to give the community visibility into node performance, stability, and reliability. The metrics may be used to influence reward calculations in future schedules.

• Forum discussion: forum.dfinity.org/t/trustworthy-node-metrics-for-useful-work/22989↗
• Background: Don’t trust, verify↗
• Technical reference: dfinity.github.io/dre/trustworthy-metrics/↗

Performance-based rewards

Performance-based Rewards (PBR) extend the schedule with incentives for keeping nodes healthy, addressing hardware and connectivity issues, and proactively reporting problems within the IC-OS stack. PBR scales each node's monthly reward by a multiplier between 0.2 and 1.0, derived from its block-maker failure rate relative to its subnet's 75th-percentile baseline.

Related

• Node Provider Documentation — the index for the role.
• Reward Configuration Guide — how to register a node for rewards once it is online.
• Proof of Useful Work — the model that frames why node providers are paid.
• Steps for Gen-1 onboarding after 48 months — receiving-provider procedure for Gen-1.1 transfers.