Fresh out of the oven
Spot a bug? These tools are new, so the odd thing may still wobble. If something looks wrong, it probably is. Tell us and we will sort it..
Fresh out of the oven
Spot a bug? These tools are new, so the odd thing may still wobble. If something looks wrong, it probably is. Tell us and we will sort it..
Fresh out of the oven
Spot a bug? These tools are new, so the odd thing may still wobble. If something looks wrong, it probably is. Tell us and we will sort it..
Pinnora's AI video generator turns a written brief into a finished clip. Choose the model, duration, resolution, audio and captions. Available lengths depend on the model; the stitching lane can join segments beyond a native take. Saved project references ground your product and presenter.
The tool is open above. This is the short version of what to do with it.
Visual style is 16 looks, from Cinematic photoreal to UGC raw phone-cam. Fill the references under it so your product and presenter are the ones on screen.
Its first line opens the render prompt, so name the moment, who is on camera and the energy right there. Up to 4000 characters.
Duration starts at 4 seconds. The selected model sets the maximum; the stitching lane can join segments beyond a native take.
Aspect ratio and model sit above these two. Captions burn from the rendered audio either way, so leaving narration empty still gets you them.
Short answers to what people ask before they start.
The Duration ruler starts at 4 seconds. Its maximum depends on the selected model and updates when you change models. The stitching lane joins segments for lengths beyond one native take. Choose an available resolution before generating; the result is an MP4 with a poster frame.
Render time depends on the selected provider, duration, queue and post-processing. The history card shows the current phase and an estimate while the job runs. Longer clips and stitching can take longer than a single native take.
When the project has an approved reference pack, the render passes your product shot, your presenter and your background in as numbered, role-named references, and the product one is told to reproduce shape, packaging, colour and label text exactly. With no approved product image it builds from the product text alone, which is the usual reason a clip shows something generic instead of your actual product.
Yes. Audio is generated with the picture and the default track is foley and room tone, where every sound has a visible on-screen cause. A music bed is opt-in. When the brief is presenter-led the voice is lip-synced natively. Captions are on by default and burned in after the render, timed from a transcript of the clip and biased with your brand terms so the name is not misspelled.
Presenter references help ground identity, but generated people can still drift between frames or segments. Use a clear presenter reference and review the finished clip before publishing. A brief that says no people excludes presenter references.
Model starts on Auto, and that row is labelled with the model it would actually pick rather than the word Auto. The dropdown lists the video catalogue with each provider and a per-second retail estimate where the rate is known, so you can pin one. Render tier is the separate dial: Draft for the fastest preview, then Standard, Premium and Veo Cinema.
The rest of the family. Every one is free to try and runs on the same engine.
