See the photo. Watch the result.
An actual Kling generation, with the input and limitations shown. No stock video presented as a customer result.
Updated 2026-10-26
New: 15-second solo, five shots
Actual output from the current Kling request: 15.04 seconds, 720 × 1280, H.264 with AAC audio, one fictional adult portrait. Per-second frames show the same red-shirt lead through the opening, laugh close-up and final band shot. This is original AI music and generated choreography, not the original Outstanding recording or an exact remake. Lip synchronization and authorized real-person inputs are not yet validated. Estimated output cost: $1.68, excluding retries and other costs.
The earlier 8-second test
Two fictional adult portrait references were used for the lead singer and keyboard player. A third sax player was invented by the model. The generated file is 8.04 seconds, 1280 × 720, with audio.
What worked
Per-second frames preserve the three roles and their clothing. A local Whisper transcription matched the requested original lyric: “We lost the game, but we kept the groove.”
What is not proven
This test does not establish real-person acceptance, precise lip synchronization, multi-shot identity continuity or Outstanding replication. A keyboard marking appeared despite the prompt requesting no logos.
Cost is not the selling price
Estimated model output cost was $0.896 for this 8-second test. This excludes retries, payment fees, storage and other processing; the invoice has not been independently reconciled.