I’ve been thinking about the current Lightning wallet landscape and I feel like there’s a big opportunity sitting right in front of us.
Right now, we basically have two extremes:
- Super simple wallets that abstract everything away (great UX), but rely on their own centralized LSP infrastructure.
- Power-user wallets that give you full control and decentralization, but are too complex for the average person.
There seems to be a missing middle ground:
A wallet that is as easy as the simplest Lightning apps — but backed by a truly open, decentralized marketplace of LSPs.
Here’s the idea:
Instead of wallets hardcoding or curating a fixed list of LSPs, we could use something like Nostr as a decentralized discovery layer.
How it could work:
- Define a standard Nostr event type for LSP announcements.
Each LSP publishes:
- Node ID
- Supported features (JIT channels, zero-conf, splicing, etc.)
- Fee model
- Min/max channel sizes
- API endpoint - Wallets subscribe to these events from known relays.
They can:
- Auto-select the “best” LSP based on fees/latency
- Or let users manually choose - The only additional infrastructure needed:
A standardized LSP API.
Something minimal like:
- Submit invoice
- Receive wrapped invoice
- Get channel quote
- Confirm channel open
Each node operator who wants to be an LSP just runs this API on top of their Lightning node and publishes their configuration via Nostr.
No central registry, no permissioned list, no gatekeeper.
- Channel opens trustlessly via wrapped invoice.
This would:
- Lower the barrier to becoming an LSP
- Encourage competition on fees and reliability
- Reduce centralization pressure
- Keep UX simple for normal users
- Preserve self-custody
It feels like most of the building blocks already exist:
- Lightning specs for JIT channels
- Wrapped invoice mechanisms
- Nostr relays and event infrastructure
- LDK-based wallet frameworks
What’s missing is just glue and coordination.
Curious what others think:
Is Nostr the right discovery layer?
What would be the biggest technical or economic challenge to making this real?
Share
Lightning.Pub is already a nostr powered Lightning node, and signals to relays with a beacon for wallets to connect. LSP services in that beacon is a goal of ours, but a longer range objective, since there's far more considerations than just advertising and an API.
Setting up a node to use nostr for comms is easy, less than a minute as we've shown, but being a good LSP is another matter entirely... just being reachable over IP4 negates a p2p-like marketplace.
Potential clients also need to know an LSP is reliable, so that implies network level insights about up-time and liquidity.
It'd be pointless to offer this for wallets specifically, since wallets are generally sales interfaces. Wallets choose LSP's not for lack of a decentralized market, but because the wallet is how they sell their LSP service. Zeus sells Olympus, Phoenix sells Acinq... and so on. LSP's are how wallets monetize.
Mobile nodes are also a retarded idea, I've articulated that many times before here.
Liquidity ads is a similar concept to yours, done over Lightning gossip, which is dumb since gossip isn't meant for that. Nostr is the appropriate forum for it and vast improvement, but Liquidity ads has no traction for many of the reasons I'm listing here.
Amboss Magma is a centralized market, it solves some of the reputation issues with all its insights as a result, but still savvy users are going to dig down into their network insights and self-peer rather than buy blindly.
Lightning Labs Pool is similar, blindly accepting an offer from anyone who was able to list isn't a recipe for success.
Lowering the barrier ultimately will just flood the market with bad LSPs, the high barrier acts as a filter of quality... the Lightning.Pub vision is less about a marketplace of arbitrary peers, but rather using your existing Nostr social graph. The best LSP is the node of someone you'd still deal with otherwise, that keeps your routes shorter and Web of Trust adds a filter beyond network insights.
The problem is not with centralization or decentralization of LSPs.
Right now you can open dozens of channels with many LSPs, whenever you want.
Try yourself to run a LN node and be the LSP for just a bunch of friends and family.
Having a good LSP node is the problem and not all node runners are capable to keep it at float.
Keep in mind that a good LSP should be also a swap service, otherwise is quite hard to manage that liquidity.
I've described such scenarios in some of my guides.
Take for example https://lightning.pub that can manage the liquidity in automated mode, with various peers.
Phoneix also is managing it in automated mode, but only with one LSP.