Add the reference
Upload the video or image in the canvas sidebar. Add product photos as well if you want the adaptation to take them into account.
Reference video analyzerAvailable now
Drop a reference video or image onto the creative canvas and get its structure laid out as editable nodes you can turn into scripts, prompts and new clips.
Open the studioSign in to create · See the credit estimate before you generate
How it works
The TrendRemix reference video analyzer turns an uploaded video or image into a structured breakdown on the creative canvas. For each shot it records the time range, camera, scene, subject appearance, action, spoken line, subtitle, audio and tags. It also plots an emotion curve across the video and writes a strategy panel that sets the source's elements beside an adaptation for your product, with a risk note. Every shot gets a generation prompt, and the whole video gets one too. You choose the response language and the language for lines and subtitles, and can generate a 4–15 second clip from any shot.
Upload the video or image in the canvas sidebar. Add product photos as well if you want the adaptation to take them into account.
Pick the languages for the analysis and for spoken lines. Use 1:1 recreate to describe the source faithfully, or edit and partial edit to apply your own instructions.
Strategy, emotion and shot nodes appear on the canvas. Edit any shot, ask the assistant follow-up questions, or generate a clip straight from a shot card.
Each shot gets its own card with time range, camera, action, line, audio, tags and notes, so the order of events is visible at a glance.
A chart of emotional intensity over time shows where the video builds, peaks and resolves, with a short reading of the curve.
Lists the source's key elements beside their equivalents for your target product, with a warning about what may not carry over.
Every shot carries a generation prompt and the video has a global prompt, so you can move from analysis to new clips or images without starting from a blank page.
| References | Video and image files from uploads or your asset library |
|---|---|
| Shot fields | Time range, camera, scene, appearance, action, line, subtitle, audio, tags, notes, prompt |
| Overview nodes | Strategy panel, emotion curve, global video prompt |
| Languages | 18 response languages; lines and subtitles set separately |
| Modes | 1:1 recreate, edit, partial edit; optional visual-only |
| Clip from a shot | 4–15 seconds, generated on the canvas |
| Cost | Credits per analysis, returned if the analysis fails |
No. It describes what can be seen and heard and suggests how the structure could be adapted. The file itself holds no information about reach, ranking or viewer behaviour.
Yes. For a still, the analysis concentrates on composition, subject, lighting and materials and produces an image recreation prompt you can send to image generation.
1:1 recreate describes the source as closely as possible. Edit treats your instructions as changes to the source, and partial edit changes only the elements you name. Visual-only leaves out dialogue, subtitles and audio.
Yes. Each shot card has its own generate control for a 4–15 second clip, and you can connect shots to image or video generation nodes on the canvas.
The analyzer gives you the breakdown and leaves the next step to you. Video remix runs analysis, planning, generation and assembly as one job and returns a finished video.
Videos you made or have permission to use. Learning from a structure is different from reusing someone's footage, voice or likeness, which needs their consent.
TAKE IT FURTHER
Decide what to keep and what to change, prepare optional product references, and review your new video against a clear brief.
Every result lands in your asset library, ready for the next tool.