drand primary sources
Reviewed: 2026-05-10 UTC
Core project framing
- https://drand.love/
- Official project site; positions drand as a distributed randomness beacon focused on verifiability, unpredictability, bias resistance, decentralization, and availability.
- https://docs.drand.love/
- Documentation entrypoint; useful for the project overview, operator docs, security model, specification, and timelock-encryption branch.
Protocol and implementation
- https://docs.drand.love/docs/overview/
- Best concise docs entry for the protocol map: cryptography, security model, specification, and operator membership path.
- https://github.com/drand/drand
- Best implementation-facing source. The README explains the DKG phase, threshold BLS signature rounds, public endpoints, and verification model in enough detail to reason about mechanism and governance.
- https://eprint.iacr.org/2016/1067.pdf
- Background academic paper linked from the repository:
Scalable Bias-Resistant Distributed Randomness.
- Background academic paper linked from the repository:
Timelock encryption branch
- https://docs.drand.love/docs/timelock-encryption/
- Best single-page source for drand’s timed-release model, use cases, and explicit security assumptions.
- https://eprint.iacr.org/2023/189
- Timelock-encryption paper linked directly from the docs.
Main takeaways from this pass
- drand is most useful as a public beacon plus threshold-governed time source, not merely as an RNG API.
- Consortium membership is a real control surface: the League of Entropy operates the production network and applicants must apply for membership.
- The threshold BLS beacon design makes the system legible as DKG + round signatures + public verification rather than opaque randomness-as-a-service.
- Timelock encryption is the most analytically distinctive extension because it turns beacon rounds into future decryption points for auctions, voting, disclosure, and anti-MEV workflows.
- The docs also surface the most important failure cases: threshold corruption can reveal future ciphertexts early, and network shutdown can leave future ciphertexts undecryptable.