SP1

  • Name: SP1
  • URL: https://docs.succinct.xyz/docs/sp1/introduction
  • Category: zkVM / proving runtime / recursive proof-system infrastructure
  • Tags: ethereum-ecosystem
  • Summary: SP1 is the zkVM inside Succinct, not a separate category anchor. Keep this note on the runtime itself: RISC-V execution, recursive proving, proof-format choices, and the curated precompile layer that makes some workloads much cheaper to prove than others. For the named implementation and release-stewardship boundary around docs, patched crates, prover defaults, and downstream distribution, use succinct-labs.
  • What it does:
    • Lets developers write provable programs in Rust and other LLVM-adjacent languages that compile to RISC-V rather than building custom circuits by hand
    • Generates different proof forms, including default STARK proofs, compressed proofs, and onchain-oriented SNARK wrappers such as Groth16 and PLONK
    • Uses recursive proving so large computations can be split into smaller chunks and then aggregated into succinct verifiable outputs
    • Ships an optimization layer of precompiles and patched cryptography crates so common operations like hashing, elliptic-curve work, and KZG-related logic become cheaper inside the zkVM
    • Provides open-source prover and verifier implementations with production-oriented documentation, examples, and audit references
  • Key claims:
    • The official docs describe SP1 as a zero-knowledge virtual machine for arbitrary RISC-V programs and say developers can use familiar languages like Rust without needing custom circuit design or deep cryptography expertise
    • The What is a zkVM? page says SP1 uses STARKs over the Baby Bear field, recursive proving for long computations, and a STARK-to-SNARK wrapping layer to make onchain verification practical
    • The proof-types docs expose a concrete verification-economics surface: compressed proofs for recursive reuse, Groth16 proofs around ~260 bytes and ~270k gas, and PLONK proofs around ~868 bytes and ~300k gas
    • The precompiles docs show that performance is not only about the VM core; it also depends on which cryptographic crates Succinct patches and supports, which is a meaningful control surface for downstream apps
    • The repo README positions SP1 as capable of proving large systems like Tendermint light clients and type-1 zkEVM-style workloads, reinforcing that it should be tracked as a distinct proving substrate rather than flattened into a generic zk tooling label
  • Whitepaper: No standalone SP1 whitepaper was surfaced during this pass. The most useful primary-source packet is the current SP1 docs plus the public repository README; see ../whitepapers/sp1-primary-sources-2026-05-12.md. The broader Succinct whitepaper remains relevant context for the parent stack but is not the cleanest source for SP1 itself.
  • Sources:

Internal linkages

  • Parent stack that commercializes and routes SP1 proofs: succinct
  • Company / implementation boundary: succinct-labs.
  • Closest broader zkVM stack to keep nearby: risc-zero

Protocol / company boundary

  • Keep projects/sp1 on the technical runtime: proof formats, recursion model, supported cryptographic precompiles, and what kinds of programs become cheap or awkward to prove inside this zkVM.
  • Use succinct-labs when the sharper question is who controls the main implementation, ships the docs and patched crate path most developers inherit, and packages SP1 inside the broader Succinct prover-network and OP-rollup stack.
  • That split matters because a zkVM can be an open technical surface while practical adoption still flows through one company-maintained implementation and documentation shell.

Control surface

  • Practical leverage sits in proof-system choices, verifier targets, recursive proving architecture, and especially the curated precompile and crate-support surface that shapes which workloads are attractive to build on SP1.

  • Keep the company boundary visible too: succinct-labs still matters for implementation cadence, default tooling, and how most downstream teams first encounter SP1 in production.

  • Last reviewed: 2026-08-12 UTC