Write the First Frame Last Frame AI Video endpoint contract

Choose endpoints that belong in the same shot rather than two unrelated hero images. Document screen direction, eyeline, lens impression, subject scale, and the editorial purpose of the final state. Inventory source rights for both images and mark any identity, product, text, or location detail that cannot change. Endpoint control is meaningful only when the transition has a defined editorial job and the two supplied states are compatible enough to review. Before delivery, preserve the reference set, and ask the delivery owner to record claim scope before the acceptance review.

Inspect the path, not only the two supplied images

Create a rack-focus handoff from a foreground object in the first frame to a background subject in the last, while retaining set geometry and camera placement. Review intermediate frames at a consistent cadence for geometry, identity, contact, lighting, camera axis, and causal order. A model can reproduce the beginning and ending while inventing an unusable bridge, so acceptance belongs to the whole shot rather than two thumbnails. Before delivery, preserve the decision history, and ask the production lead to record reversal cost before the rights check.

• Opening-state fidelity For this decision, preserve the decision history, and ask the source custodian to record reversal cost before the reversible handoff.
• Causal intermediate motion For this decision, preserve the handoff draft, and ask the production lead to record source fidelity before the dated decision.
• Closing-state and edit usability For this decision, preserve the input snapshot, and ask the brand reviewer to record identity consent before the dated decision.

Separate capability, availability, and entitlement

A visually attractive bridge still fails when focus travel lacks motivation, the subject jumps depth, intermediate frames warp, or the ending cannot cut cleanly into the next shot. Verify the exact vendor, model version, region, account surface, input constraints, and check date before relying on the feature. The documented Veo example proves a named control in that context; it does not prove current SEELE access or any free, downloadable, permanent, or open-ended use. For this decision, preserve the review copy, and ask the delivery owner to record failure conditions before the reversible handoff.

Keep the First Frame Last Frame AI Video evidence ledger narrow

Google DeepMind first-party documentation verifies a current first-and-last-frame control for Veo 3.1. Record the source owner, product or model name, verification date, surface or region, exact claim, and unresolved gap beside the test. Taxonomy eligibility and search demand explain why the page exists, but neither can be reused as proof that SEELE or another provider exposes the entire query as a current capability. For a controlled test, preserve the delivery checklist, and ask the rights reviewer to record destination fit before the scope confirmation.

• Google DeepMind lists first and last frame among the documented creative controls for Veo 3.1. For a controlled test, preserve the evidence table, and ask the creative lead to record claim scope before the scope confirmation.