On April 29, we’ll quietly open the doors to Spark’s beta program. For the next six months, participants will be among the first to explore the network, test its capabilities, and help shape its direction. This is an opportunity to work closely with the core team as Spark evolves.

Spark’s beta marks our first public mainnet release — the moment developers can finally get hands-on with Spark’s SDKs and start building.

Things are still early, so you may run into rough edges or limitations. Spark is fully open source and under active development — we welcome all feedback, questions, and contributions. No contribution is too small; every bit of input helps shape the network.

​
What should I expect during the beta?

Spark is launching as a mainnet beta — which means it’s real, but still early. We’re excited for you to try it out and help shape what comes next. That said, Spark is still experimental infrastructure, so we want to set clear expectations:

  • Start small — Please use modest amounts while testing. This helps us all learn safely and improve the experience together.

  • Limited operator set — Spark currently runs with just two Spark Operators (SOs): Lightspark and Flashnet. We’ve intentionally kept the set small to simplify debugging and testing during the early phase — more operators will be added soon as we scale.

  • Spark is designed to be incredibly fast and scalable — but in this early version, we haven’t yet optimized for performance across the board. Our priority so far has been security and correctness. Expect rapid improvements in speed, throughput, and UX in the short to medium term as we scale the network.

  • Unilateral exit isn’t live (yet). We know this one’s important — and yes, we wish it were ready too. The good news is: it’s coming very soon. Unilateral exit support is at the top of our list and will be enabled in the next few weeks.

  • Our SDKs connect to the Lightspark SSP by default, though you’re free to integrate with another SSP of your choice. While we don’t anticipate issues with Lightspark’s service, we may apply limits to swap amounts — particularly between Lightning and Spark — during periods of high network activity.

We’re going to be headsdown focused on shipping the critical pieces, scaling the network, and distributing it with more partners. We don’t expect everything to be perfect on day one. We’ll iterate fast and update aggressively.

We’re actively monitoring the network — your feedback during this beta is incredibly helpful. Let us know what breaks, what works, and what surprises you.

​
What SDKs can I use to build on Spark?

We’re rolling out SDKs quickly and will prioritize what you need. If there’s a language or platform you’d like supported, just let us know.

Spark Wallet SDK

For developers building wallets on Spark

  • Supported now: (TypeScript)

  • Coming soon: , , + more based on demand

Spark Issue SDK

For issuers to mint, burn, and manage assets (like stablecoins) natively on Spark.

  • Supported now: (TypeScript)
  • Coming soon: + more based on demand

​
What are we focusing on in Q2?

We have a packed roadmap ahead. Our top priority is making our infrastructure rock solid, solving all edge cases, and adding more SDKs. Please let us know if there’s anything we might be overlooking.

  • Unilateral exits are not yet ready but are coming soon with built-in watchtowers.
  • L1 support for tokens is coming soon.
  • Spark Explorer coming out
  • Improved edge case and error handling to automatically recover.
  • Improved stability/reliability.
  • Adding more SOs with threshold signing.
  • Multisig/MPC support.
  • Atomic swaps between tokens and sats.
  • 0 sat Lightning invoice support.
  • Single-use Spark address support / Spark payment requests.
  • Increased performance.
  • 0 downtime node upgrades.
  • Rust, Flutter, Swift SDK