Pond5 dataset earnings for footage contributors

Pond5’s contributor materials distinguish marketplace licensing from Dataset Earnings. The published payout guide describes a dataset program, says contributors may opt out, and separates the reporting of dataset transactions from ordinary marketplace sales. Its policy for whether contributors may submit AI-generated stock is a different question. Fiund also licenses training footage. Pond5 Payout Overview · AI-generated content submission policy

Check rights at the clip level

A catalog of footage may mix outright owned work, clips commissioned by a client, talent performances and background music. Map each clip to ownership, releases and any exclusivity program. Ask Pond5 whether a specific clip is eligible for datasets and how an opt-out affects future use. Review what the account displays for ordinary earnings, partner earnings and dataset earnings before reconciling amounts. The written agreement governs existing licenses. Pond5 Contributor Agreement

What to check if you are sourcing footage for training

A downloadable stock clip does not itself grant the right to train a model. If you need training data, ask for a written license and a description of the assets, technical specs and permitted model uses. Buyers who need the same physical task recorded from multiple angles may need commissioned capture. Rights holders who have licensed files elsewhere should reconcile those rights before offering a new package.

Who it fits. Stock video producers can review Pond5’s dataset participation options alongside their ordinary licensing activity. Start with the current agreement and your actual account settings. An account in good standing gives no guarantee of dataset inclusion or a paid transaction.

Read the published dataset share and the actual statement

The payout overview checked November 5, 2026 states a 20% royalty share for each item licensed through Dataset Earnings. It also explains that these distributions are less regular than ordinary marketplace activity and that an opt-out concerns future datasets. An account can still receive a distribution connected to an earlier deal.

The percentage does not supply the underlying transaction price or establish that a particular clip will be selected. Reconcile the actual Dataset Earnings line with its reporting period and receipt. Keep ordinary stock royalties and training-related distributions separate when comparing the return from a portfolio.

An original three-clip inventory

Imagine CLIP-A is a controlled product shot, CLIP-B is a client-owned event recording and CLIP-C contains an original performance with licensed background music. Record the source master, creator authority, releases and excluded components for each. A contributor account containing all three does not make their permission scopes identical.

For a separate training opportunity, ask which files and components are requested and what previous grants constrain them. A buyer should obtain the actual training licence and specimen rights evidence for the selected assets. A creator should review account settings, representation and the agreement that applies to the particular portfolio before offering it elsewhere.