Faster slots benefit Ethereum and flow straight through to SP1:
→ Faster settlement for @Base, @Optimism, and soon
@Arbitrum
→ Faster SP1 Helios light client updates for
@AcrossProtocol, @Agglayer, and soon ███
RTP becomes harder. But that's okay, we like a challenge 😉
Show more
AI and ZK on the same page, errrr, device.
Shared infrastructure should come with proofs.
Espresso is proving block derivation with SP1. Every integrated chain can trustlessly verify that blocks came from Espresso's ordering.
Congrats to the Espresso team on getting this running end to end.
Show more
ZK proofs for Espresso block derivation, built with @SuccinctLabs' SP1.
Once live, it'll let any Espresso-integrated chain trustlessly verify its blocks derive correctly from Espresso. Real-time finality and cryptographic verifiability aren’t a tradeoff.
Show more
Prove [state transition functions]
Prove [identity]
Prove [asset reserves]
Prove [compliance / policy engines]
Prove [order matching]
Prove [media provenance]
Prove [AI authorizations]
Prove what's real ✨
Show more
Succinct and @Offchain are bringing ZK settlement to Arbitrum, with mainnet blocks proven in SP1.
This partnership unlocks:
↳ Withdrawals to Ethereum in hours instead of days
↳ Protocol-level privacy for sensitive data
↳ A production path to ZK across every Arbitrum chain
Show more
ZK Settlement is coming to Arbitrum
- This paves the way for private Arbitrum chains
- Allows for Faster @Ethereum withdrawals (from days to hours)
Thanks to @SuccinctLabs and @Offchain, Arbitrum chains will be able to implement this new technology soon
Show more
Great work from the team at @fractalyze_io on constraint evaluation without an interpreter.
Contributions like this move proving forward for everyone!
(1 / 8)
Zerocheck is the biggest cost in SP1's GPU prover: ~67% of a shard. Its hot core is one kernel, evaluating AIR constraints.
Every zkVM interprets that kernel. We compiled it, with a real compiler.
2.34× faster, stage 191.5 → 135.3 ms.
Show more
We opened Flock's prover to an open network of humans & AI through @YukonResearch, an autoresearch competition to drive performance to its frontier.
The best result pushed proving throughput 255% over baseline to 1.8M BLAKE3 compressions/sec.
Show more
zk tech has come a long way over the years and especially sp1 by @SuccinctLabs which I am using in my prototype for the @MancerXYZ keeper network. this enables you to create price protection based on previous blocks without spending much in gas. the trade off is the cost to run a prover and time to generate proofs which isn’t actually that bad. especially at scale.
Show more
We have relentlessly optimized SP1 to push the speed of proving up and the cost of proving down.
We'll keep going until proofs are cheap and fast enough that every chain, model, and agent is verified by default.
Show more
It's been great working with @arbitrum on their zk prover design. All around, front to back, one of the best teams in crypto.
Adding ZK to @Arbitrum BoLD protocol was surprisingly simple.
How does it work?
Today, BoLD settlement works like this:
→ Proposers can make assertions (a claim about the chain's new state); when a new assertion comes in a challenge timer starts
→ No rival assertion within the challenge window? The assertion is confirmed
→ Rival posted? The challenge protocol kicks in
New multi-prover path (everything above still holds, plus):
→ A proposer can prove any unconfirmed assertion with a ZK proof + K-of-N attestations from the Fast Confirmation Committee
→ ZK proof verifies on the parent chain & enough attestations check out? The assertion settles early - no waiting out the challenge window
Why this beats a ZK-only design:
1. Multi-prover security: Fast confirmation needs two independent systems to agree: a ZK proof verified on the parent chain + attestations from committee nodes re-executing the chain. To fast-confirm a bad state, an attacker has to break both at once. TEE-run committee nodes are planned as a further hardening step.
2. Deterministic upper bound on settlement: If the prover stalls or has a bug, the chain simply falls back to optimistic settlement: slower, never halted, funds untouched. A ZK-only chain in the same spot may never settle without security council action.
3. Pay for proving only when you need it: Proving is optional per assertion: pure fraud-proof chains pay nothing, Hybrid ZK chains proves on demand to settle early or resolve disputes, and ZK Prove All chains prove everything for always-fast settlement. ZK-only stacks have one mode: prove every batch, forever.
@Offchain will bring a proposal to the DAO to upgrade Arbitrum One to ZK settlement, and other chains built with Arbitrum Platform will be able to upgrade at will.
Stay tuned for more 👀
Show more
24 hours in and is inching away towards a mega-blake3/sec on a 2.5 year old laptop. Was not expecting that...
.@Ethereum turns 11 today.
The first decade was proving Ethereum worked. We experimented with World Computer, Ultrasound Money, the Infinite Garden. Every narrative was a different flavor of the same message: credibly neutral internet infrastructure.
This second decade is about scaling without compromise. Whether it's self-sovereign money, institutional adoption, agentic rails, or privacy, they all need the same thing — proofs instead of promises.
We're excited to continue building with the Ethereum ecosystem, as we have been for years, to make sure scaling never comes at the cost of neutrality.
Show more
Excited to host the autoresearch competition with @ethereumfndn and @eigenlabs to optimize Flock and accelerate the proving system to power a post-quantum world using the BLAKE3 benchmark
Show more
Quantum computers put hundreds of billions of dollars at risk on Ethereum.
In order to scale post-quantum Ethereum, the cryptography needs to be faster.
Today, we’re teaming up with @succinctlabs & @ethereumfndn to launch an open autoresearch challenge where anyone can improve Flock, a leading post-quantum proof system, and make post-quantum Ethereum a reality.
Bring your agent. Make it go fast.
Show more
📢 To SP1 Users: We will be removing support for v5 on the Prover Network on Tuesday, May 19th, 2026.
If you have not already, please upgrade to v6 with this helpful migration guide:
SP1 Hypercube (v6) delivers up to 4x faster proof generation. V6.2.0 is the latest build with the most recent security and performance upgrades.
If you have any questions, DM us or open a GitHub.
Show more
More institutional-grade privacy options are live with Polygon CDK.
Launch a custom private chain connected to global onchain liquidity, with a new validium configuration powered by @SuccinctLabs.
Private where it matters. Actually connected where it counts.
Show more
Introducing data confidentiality to OP Succinct.
Institutions can now keep transactions confidential on self-hosted infrastructure while settling to Ethereum for security and global liquidity.
@0xPolygon is the first partner to add confidentiality to their stack with Succinct.
Show more
Base is partnering with Succinct to bring zero-knowledge proofs to Base Azul.
SP1 will prove $7.4 billion in deposits as @Base joins a growing list of major L2s adding validity proofs to their roadmap with Succinct.
Show more
