Home · FAQ
Frequently asked questions
Everything teams ask before running a node or wiring verdicts into a trust & safety pipeline. Every answer traces to the repo — no benchmark theatre, no implied partnerships.
The basics
+ What problem is dfpn actually solving?
Synthetic media — face swaps, voice clones, AI-generated images and video — is getting cheaper to produce and harder to spot. Centralized detection services concentrate that authority in a few companies; they create single points of failure and an opaque definition of "real." dfpn decentralizes detection so no single party owns the verdict.
+ Is dfpn a deepfake detection model?
No. dfpn is a coordination layer. Detection models are run by independent operators using their own hardware. The reference worker client ships with four pre-configured models covering face manipulation, AI-generated images, video authenticity, and voice cloning, but operators and model developers can register more.
+ What media types can dfpn evaluate?
The reference worker covers image, video, and audio — including face manipulation and voice cloning. The model registry is permissionless, so coverage grows as developers register new detectors.
+ What is dfpn not trying to do?
It is not a takedown service, a legal-attribution mechanism, or a private on-chain inference platform. Those are out of scope by design. dfpn answers a narrower question: "does the network agree this piece of media is manipulated, and on what evidence?"
Verifiable AI & DePIN
+ How is dfpn an example of verifiable AI?
Rather than trusting one model's output, dfpn has many independent operators run detection, commit-reveal their results so they cannot copy each other, and produce a reputation-weighted consensus with outliers slashed. The verdict is a staked, auditable agreement — a concrete, narrow instance of verifiable AI, not a general-purpose "AI oracle."
+ Why is dfpn described as a DePIN network?
Detection is GPU-bound physical work. dfpn coordinates a decentralized physical infrastructure of operator GPUs: contributors bring real hardware, earn DFPN for accurate inference, and network capacity scales with whoever plugs in — no central data center and no single kill switch.
+ Does dfpn verify arbitrary AI outputs, or just media?
Just media authenticity, deliberately. The output is a single checkable claim — manipulated, authentic, or inconclusive — which is exactly what makes staked consensus work. dfpn is not attempting to be a general verifiable-inference or "AI oracle" layer, and it stays honest about that scope.
How it works
+ Where does inference actually run?
Entirely off-chain, on operator hardware. The on-chain programs track request lifecycle, commitments, reveals, scoring, and payouts. The chain holds hashes and metadata, not media.
+ How is collusion prevented?
Commit-reveal stops one worker from copying another. Random assignment, diversity constraints, reputation weighting, and a challenge window before slashing make coordinated attacks costly. The full threat model is in the dfpn docs.
+ How is a verdict actually computed?
Several independent workers each run their own models and commit a hashed result, then reveal it. The chain aggregates the revealed results with reputation weighting into a verdict label and a confidence score. It is not a naive majority vote.
+ Can a verdict be rewritten after the fact?
No. Every commitment, reveal, and aggregated verdict is anchored on Solana. There is no admin endpoint that silently flips a result — the audit trail is reconstructable by anyone.
Economics & participation
+ How does economic security actually work here?
Workers stake DFPN before they can serve requests. They earn a share of fees when their results agree with consensus, and they get slashed for invalid results, fraud, or missed deadlines. Model developers stake separately and earn a cut whenever their model is used.
+ What are the stake floors?
Workers post at least 5,000 DFPN; model developers post at least 20,000 DFPN per model version; challengers escrow 5% of the disputed reward. Floors scale with request fees per epoch.
+ How are fees split?
Per request: workers 65%, model developers 20%, treasury 10%, insurance pool 5%. Fees are paid in SOL by default (minimum 0.02 SOL) with an optional DFPN discount.
+ How do I earn as an operator?
Stake DFPN, run the worker daemon on your own GPUs, and process requests accurately and on time. Rewards weight accuracy 50%, availability 25%, latency 15%, and consistency 10%. See the run-a-node quickstart to get started.
Status & positioning
+ Is this production-ready today?
The README is upfront about the timeline: devnet, then testnet, then mainnet beta. Read the milestone plan and roadmap in the repo before integrating, and treat current numbers as forward-looking until mainnet beta.
+ How does dfpn relate to provenance standards like C2PA?
They are complementary. Provenance answers "where did this come from?" via signed manifests at capture time; detection answers "does this look generated?" after the fact. Use both: provenance when producers cooperate, dfpn when they did not. See the C2PA and Truepic comparisons.
+ How does dfpn compare to centralized detection vendors?
Vendors like Reality Defender, Hive, and Sensity offer strong first-party APIs but concentrate the verdict in one company. dfpn distributes it across independent operators with an on-chain audit trail. Compare them directly on the comparison pages.
Still have a question?
Talk to the team about running a node, publishing a model, or integrating the verification flow.