Pyth Entropy primary sources
Reviewed: 2026-05-15 UTC
Official overview / positioning
-
https://docs.pyth.network/entropy
- Best short official framing in this pass.
- Used to confirm that Pyth Entropy is positioned as an onchain RNG for developers, with commit-reveal-based verifiability, low-latency delivery, and v2-era integration improvements.
-
https://docs.pyth.network/llms-entropy.txt
- Useful curated quick-start because it consolidates current terminology around request/response flow, callback delivery, fee lookup, provider defaults, reveal delay, and the Fortuna provider API.
- Especially helpful for current operational framing even though it is a summary layer rather than the most canonical mechanism source.
Core mechanism / trust assumptions
- https://docs.pyth.network/entropy/protocol-design
- Best mechanism source in this pass.
- Used to confirm the two-party commit-reveal structure, provider-side hash-chain precommitment, request sequence numbering, later reveal verification via hash-chain linkage, and callback-based delivery to the consuming app.
- Most important source for the explicit trust caveats: provider censorship after seeing outcomes, front-running or collusion risk, and the danger if the provider keeps or leaks the hash chain.
v2 product surface / operational layer
-
https://docs.pyth.network/entropy/whats-new-entropyv2
- Best concise source for how the current product surface differs from older Entropy materials.
- Used to confirm multiple request variants, callback-status tracking, and the Entropy Explorer / retry-oriented operational layer.
-
https://docs.pyth.network/entropy/request-callback-variants
- Best source for the request-surface decomposition.
- Used to confirm that Entropy v2 explicitly separates default request, custom gas-limit request, custom-provider request, and full-control request with explicit user entropy.
-
https://docs.pyth.network/entropy/set-custom-gas-limits
- Best source for showing that callback gas budgeting is not incidental but part of the protocol’s fee and liveness envelope.
- Used to confirm gas-limit rounding, minimum/default-provider behavior, maximum bound, fee dependence on callback gas, and the way callback complexity feeds into integrator cost and reliability.
-
https://docs.pyth.network/entropy/debug-callback-failures
- Best source for the manual recovery / operator-debug path.
- Used to confirm
revealWithCallback(), event-log-derived request debugging, provider-contribution retrieval, reveal-delay dependence, and the fact that callback failure recovery is an expected operational workflow.
Interface / SDK-level trust split
-
- Best source for the interface-level security nuance.
- Used to confirm that the simplest
requestV2()overloads use an in-contract PRNG for the user contribution, explicitly changing the trust assumption so a dishonest validator-plus-provider pair becomes the relevant manipulation risk. - Also used to confirm fee lookup functions, default-provider lookup, provider-info queries, request-state queries, and gas-limit behavior directly from the interface comments.
-
- Useful implementation-oriented source.
- Used to confirm the older explicit commit / request / fetch-provider-number / reveal flow, the native-token fee pattern, and the Fortuna-style provider API path that sits beside the onchain verification logic.
Example integrations
- https://github.com/pyth-network/pyth-examples/tree/main/entropy/coin_flip
- Useful for grounding how Entropy is consumed by ordinary app contracts and clients.
- Used to confirm that the current example path is callback-based and sequence-number aware, and that Pyth itself distributes example apps for common RNG-style integrations.
Main takeaways from this pass
- Pyth Entropy is best understood as provider-mediated, callback-priced, request-response randomness middleware rather than a generic oracle bullet under the broader Pyth brand.
- The most analytically useful split is between the cryptographic core (hash-chain commit/reveal), the developer-facing request surface (provider choice, gas choice, explicit or implicit user entropy), and the operational layer (callback liveness, explorer/debug flows, manual recovery).
- Entropy clears the corpus bar because it makes visible a family of control surfaces that many randomness products flatten away: reveal liveness, provider choice, callback gas pricing, overload-specific trust assumptions, and the offchain service path that helps requests actually complete.