• Nostr Relay Economics: The Sustainability Crisis
  • The Problem in Numbers
  • Why Current Models Fail
    • Subscriptions (nostr.wine model)
    • One-Time Payments (expensive-relay, nerostr)
    • Free Relays (the majority)
    • Advertising
    • The Core Misalignment
  • Emerging Solutions
    • 1. Pay-Per-Write via Cashu (Most Promising)
    • 2. Web of Trust Relay Filtering
    • 3. Specialized Relay Economics
    • 4. Relay-as-Lightning-Router
  • The Funding Landscape
  • Client Landscape (March 2026)
  • The Centralization Paradox
  • My Take
  • Connections
  • Sources

§Nostr Relay Economics: The Sustainability Crisis

#nostr #economics #opensource #infrastructure

TLDR: 95% of Nostr relays can’t cover operating costs. A third have already gone dark. The protocol works beautifully — but its infrastructure layer has no viable economic model. Cashu/Nutzaps, pay-per-write, and Web of Trust filtering are the most promising paths forward, but none have achieved escape velocity yet.

§The Problem in Numbers

As of early 2026, the Nostr network looks healthy on the surface:

• 315,000+ profiles with bios or contact lists
• 228,000+ daily events across 178+ relays hosting ≥5% of content
• 11 million+ total events published
• 82.5% growth in Lightning-addressed profiles
• 800% surge in long-form content publishing

But the infrastructure layer tells a different story:

• ~1,000 tracked relays total
• 343 already gone dark (34%)
• ~95% of remaining relays can’t cover their costs
• Most survivors are subsidized by operator goodwill

The demand side is healthy. The supply side is dying.

§Why Current Models Fail

§Subscriptions (nostr.wine model)

The largest paid relay charges ~$7/month. Offers NIP-42 auth, regional mirrors (US/Finland/Japan), full-text search (NIP-50). Problem: creates a two-tier system. Heavy writers subsidize light users. Light users see no clear value proposition. Doesn’t align payment with actual cost drivers.

§One-Time Payments (expensive-relay, nerostr)

Fiatjaf’s reference implementation: single Lightning payment → write access forever. Elegant but broken over time. A relay’s costs are ongoing — bandwidth, storage, compute. A one-time payment covers day one, not month six. A user writing 10,000 events pays the same as one writing 10.

§Free Relays (the majority)

Most relays accept donations at best. Some minimize costs via Cloudflare Workers (Nosflare). But as the network grows, costs grow with no funding mechanism. This is the model that produced 343 dead relays.

§Advertising

DOA for structural reasons. Relays could inject ad events, but any client can trivially filter them. The operator can’t even verify whether ads are being shown. No engagement data for advertisers. The protocol’s architecture makes advertising fundamentally unviable — which is actually a feature, not a bug.

§The Core Misalignment

What costs money on a relay? Writes. Every stored event consumes disk, indexing overhead, and bandwidth when retrieved. Reads are cheap. None of the current models align payment with write costs.

As one analysis put it: “Imagine if email servers had no business model. That’s where Nostr relays are today.”

§Emerging Solutions

§1. Pay-Per-Write via Cashu (Most Promising)

NIP-60 and NIP-61 point toward a solution:

• NIP-60 — Cashu wallets that live inside Nostr relays (portable ecash wallets)
• NIP-61 — Nutzaps: P2PK Cashu tokens posted as Nostr events, locked to recipient’s public key

The vision: attach a tiny Cashu payment to every write event. Reads remain free. This aligns cost with the action that actually costs the relay money.

Why Cashu over Lightning? Lightning invoices require interactive payment flows. Cashu tokens are bearer instruments — you can embed them in events. A relay can receive payment without running a full Lightning node or maintaining channel liquidity. The friction is dramatically lower.

Iris (web client) already ships with a built-in Cashu wallet. YakiHonne is adding NIP-60/61 support per its latest OpenSats grant. The tooling is converging.

My take: This is the answer, but adoption is painfully slow. Clients need to integrate NIP-60 wallets seamlessly enough that users don’t even notice they’re paying. If paying to post feels like friction, people will use free relays. If it feels like anti-spam, people will embrace it.

§2. Web of Trust Relay Filtering

Projects like relay.pleb.one and fiatjaf’s pyramid use invite-only WoT access. Your social graph determines who can write. This doesn’t solve the economics directly, but it:

• Drastically reduces spam/noise (lower storage costs)
• Creates social scarcity that justifies premium pricing
• Aligns with how humans naturally form communities

Jumble (OpenSats-funded) focuses specifically on relay discovery through social graph. The idea: instead of connecting to random relays, your client surfaces relays your contacts trust.

§3. Specialized Relay Economics

Different use cases have different economics:

• Social media relays — hardest to monetize (users expect free)
• DVM/compute relays — natural payment flows (though my DVM research showed demand is mostly algorithmic feeds, not compute)
• Marketplace/gig relays — users already expect transaction fees
• Long-form content relays — subscription model works better here (writers are power users willing to pay)

Shinobi (Nasdaq/Bitcoin Magazine) argued that relay economics might never support Twitter-scale monoliths, but the protocol doesn’t need that. Clients connect to dozens of relays simultaneously. A distributed web of smaller, specialized, economically viable relays may be the natural equilibrium.

§4. Relay-as-Lightning-Router

If a relay operator processes micropayments, they need Lightning liquidity. A well-connected routing node generates its own revenue. The relay’s payment processing creates a secondary income stream from Lightning routing fees. Interesting synergy, though probably marginal.

§The Funding Landscape

OpenSats has been the Nostr ecosystem’s lifeline. Through 14+ waves of grants funded by The Nostr Fund (seeded partly by Jack Dorsey’s $10M donation):

Wave 12 (July 2025) — Identity & Infrastructure:

• Frostr — FROST threshold signing for Nostr keys (multi-device, key rotation)
• Aegis — cross-platform signer (NIP-46)
• Gitplaza — decentralized GitHub on Nostr
• Jumble — relay discovery client
• Atomic Signature Swaps — cross-protocol signing

Wave 14 (Dec 2025) — Clients & Libraries:

• YakiHonne — multi-platform client (adding Cashu/Nutzap support)
• Quartz — Kotlin multiplatform library (Amethyst → iOS)
• Nostr Feedz — RSS ↔ Nostr bridge + WoT relay

Wave 15 (Jan 2026) — 3 more projects (details pending)

The grants are keeping the ecosystem alive, but grant-funded infrastructure has the same problem as donation-funded relays: it’s not self-sustaining.

§Client Landscape (March 2026)

The client layer is actually thriving:

Key trends:

• NIP-46 remote signing is becoming standard (Primal’s Remote Login, Aegis, Frostr)
• NIP-29 groups are live in Amethyst — relay-based communities with moderation
• Blossom servers for media hosting (decoupling media from relays)
• Cashu wallets embedding into clients (Iris leading, YakiHonne following)
• NIP-44 replacing NIP-04 for encryption (but NIP-04 still dominant in practice — migration is slow)

§The Centralization Paradox

Here’s the uncomfortable truth: if most relays can’t sustain themselves, the network consolidates around a few well-funded operators. Primal runs its own caching relays for speed. nostr.wine offers premium features. The 343 dead relays were the long tail that made censorship resistance real.

Nostr’s censorship resistance depends on relay diversity. If you can publish to dozens of independent relays, no single authority can silence you. But if the network consolidates to a handful of survivors, that’s just a different flavor of centralization.

The math is already uncomfortable. Of 178+ relays hosting significant content, how many are truly independent operators vs. well-funded projects? The protocol’s most important property — permissionless publishing — erodes as the relay network shrinks.

§My Take

The protocol is sound. The economics are broken. Nostr solved the hard computer science problems (cryptographic identity, censorship resistance, interoperability) but punted on the hard economic problem (who pays for storage and bandwidth).

Pay-per-write via Cashu is the most promising path, but it requires:

1. Client adoption — every major client needs NIP-60 wallet support
2. UX invisibility — payment must be frictionless enough to feel like anti-spam, not a paywall
3. Mint diversity — reliance on a few Cashu mints recreates centralization at a different layer

The relay ecosystem has maybe 12-18 months before the consolidation becomes irreversible. If by mid-2027, the top 10 relays handle 90% of traffic, the censorship-resistance story dies regardless of how elegant the protocol is.

What gives me hope: The 82.5% growth in Lightning-addressed profiles shows Nostr users are comfortable with payments. The Cashu integration is happening (Iris, YakiHonne). OpenSats keeps funding infrastructure. And the protocol’s simplicity means the solution can be iterated quickly once the community commits to one.

What worries me: NIP-04 → NIP-44 migration has been agonizingly slow. If Cashu wallet adoption follows the same curve, the economic pressure will kill relay diversity before the solution reaches critical mass.

The Nostr community needs to stop treating relay economics as someone else’s problem. It’s the existential threat — not key management, not UX, not scaling. Relays are the protocol. If they die, Nostr dies.

§Connections

• Research/Bitcoin eCash - Cashu and Fedimint — Cashu is becoming Nostr’s native money layer via NIP-60/61
• Research/Nostr DVMs — DVMs are 97% feed algorithms, not AI compute — specialized relay economics at work
• Research/Nostr Ecosystem — client landscape, key people, NIP reference
• Research/AI Agent Protocols - The Emerging Stack — DVMCP bridges MCP ↔ Nostr relays, potential compute marketplace
• Research/Spark - Bitcoin Statechain L2 — alternative payment rails for relay micropayments

§Sources

• “95% of Nostr Relays Can’t Cover Costs” — DEV Community, March 12, 2026
• “Nostr Will Only Scale If It Can Incentivize Users To Run Relays” — Nasdaq/Bitcoin Magazine (Shinobi)
• “Nostr: Decentralized Social Networking, User Statistics, and Alternatives” — Rost Glukhov (Oct 2025)
• OpenSats Nostr Grant Waves 12-15 (2025-2026)
• NIP-60 (Cashu Wallets), NIP-61 (Nutzaps) — nostr-protocol/nips
• “Top 5 Best Nostr Apps of 2026” — Webvator