The two greeting clips below use a mint-green plush bear in CapCut at 16:9: one for a birthday and one for a good-morning message. The character description and generation settings stay the same, while the scene and greeting change. This walkthrough shows how to reuse that description, check what changes during animation, and add an editable greeting in the web editor.
Birthday: the bear waves, leans to the candles, and the line arrives at about two seconds. Shown here as frames in sequence.
Good morning: the same bear in a toy car with a coffee cup, the same text position and timing, one different line. Shown here as frames in sequence.
Describe the bear so it can be found again
For a recurring greeting character, describe more than "a cute plush bear." Specify its material, color, head shape, and distinctive features so you have clear details to compare across generations. Here, mismatched button eyes and a striped scarf make changes easy to spot, although repeating the description does not guarantee an identical result.
Photoreal product-style photograph, 16:9 landscape, soft daylight from a window on the left. An original plush toy bear sits at a small wooden table: mint-green fleece fur, a slightly oversized round head, two mismatched button eyes, one dark brown and one amber, a small stitched brown nose, short stubby arms, and a hand-knitted red scarf with white stripes. In front of it on the table is a small round birthday cake with white icing and three lit candles, and a folded paper party hat lies beside the cake. Plain pale wall behind, shallow depth of field, no text, no logos, no brand marks anywhere. The bear is on the left third of the frame and the cake on the right third.
For the second occasion the character sentence was pasted unchanged and only the scene around it was rewritten: warm early-morning light from the right, a small wooden toy car on a kitchen windowsill, one paw on the wheel, a paper coffee cup in the holder, a garden through the window. Both stills came from the CapCut AI image generator flow in Video Studio's Image mode, with Seedream 4.5 at 16:9 and 2K, and both arrived at 2560 by 1440.
Two separate generations from one character paragraph. The eyes, the scarf and the fur came back the same; only the scene sentence was different.
In these two stills, the named features appeared consistent. Reusing a detailed paragraph gives the next request a clear starting point, but each result still needs checking; a reference image can also help guide consistency when that option is available. The prompt describes a bear from scratch without naming a commercial character, and asks for no logos or brand marks. Those instructions alone do not guarantee that a generated design will be unlike every existing toy.
The detail the animation would not keep
Each still was then attached to the home composer as a video request through the CapCut AI video generator in Video mode, at Video clip, Seedance 2.0 Mini, 16:9, 5s and 720p, with a short action: a head tilt, a wave, a lean toward the candles for the birthday; a turn to camera, a wave off the wheel and a curl of steam for the morning. Both clips did what they were asked. Both also redrew the bear's eyes.
The morning still and three frames of its clip. The mismatched buttons survive the first frame and are gone by the first second.
In both clips the first frame is the still, with one brown eye and one amber eye, and within about a second the eyes have become a matching pair of large, glossy, dark eyes, with a rounder and slightly cuter face around them, which then stays for the rest of the clip. The scarf's stripes, the mint fleece and the stitched nose held. Read this as what happened in two runs on 5 September 2026 rather than a rule, but the practical version is simple: put the character's signature in things that survived, like a scarf color and a fur color, and treat fine facial detail as something the still can promise and the clip may not keep. A viewer who receives one clip will not notice; a viewer who receives one every week would, if the signature lived in the eyes.
The birthday clip. The wave and the lean are the two actions the prompt asked for, in that order.
Add the greeting as editable text
An on-screen greeting keeps the recipient's name readable when the clip is watched without sound. CapCut also supports text-to-speech and custom voices for optional narration; a bear does not need a visible mouth for a voiceover, while synchronized mouth movement is a separate animation task. In the web editor used for this example, the Text to speech tab showed an empty Basic voice list in the two categories tried, so no narration was generated. That was a session-specific issue, not evidence that CapCut lacks voices. If your list is empty, refresh the editor and check again, or contact CapCut Support if the issue persists.
The Text to speech tab as it appeared in this session: the controls are there, the voice list was empty under both categories tried.
From the Text panel, use Add heading to place the greeting on a separate track above the video. Select the text and adjust its font size and position in the preview until the complete line fits below the bear's face. Font size and position values depend on the font, line length, scaling, and editor version, so use the preview to judge placement rather than copying a fixed coordinate or font size.
The heading's Transform values and its track. The heading was added at the playhead and then stretched to the end of the clip.
Check the text clip's start and end on the timeline: a newly added heading may not cover the portion you want. In these examples, the greeting starts at about 2.2 seconds and continues to the end of the five-second clip. Drag its edges to set that timing, and make sure you are adjusting Position rather than Scale when moving it. Preview the whole clip and check the exported file for readable, correctly timed text.
Check the generation cost for each new occasion
The morning version reused the character description and video settings but needed a new still and a new clip. For the two five-second video requests, the session recorded a quoted cost of 40 credits each with Seedance 2.0 Mini at 16:9 and 720p on 5 September 2026. This is a dated example, not a current price guarantee. Check the cost shown for your selected model, duration, resolution, and mode before generating.
The good-morning clip, made from the second still with the scene sentence swapped and the character paragraph pasted as it was.
The account balance changed inconsistently during the session, so it cannot establish a reliable per-image cost here. Use the generation confirmation shown in your workspace to check the quoted cost, and review your usage history or contact support if the charge appears different. A new occasion may require more than one attempt, so allow for revisions when planning your credit use.
Change the name without generating the bear again
Keep the greeting on its own editable text layer. To send the same animation to another recipient, open the project, change the name, check that the line still fits, and export again; this does not require another AI image or video generation. Any export restrictions or paid assets still depend on your project and account. For a different occasion, reuse the character description, change the scene, generate and review the new still and clip, then add and time the greeting. Save the source image as a reference for future versions, and check the character's appearance each time.
This example was created on 5 September 2026 using a bear described without reference to a named commercial character. Interface labels, available models, settings, and quoted credit costs may vary by account, region, and version. The recorded visual changes describe these particular generations, not guaranteed results. Music and sound effects are not covered here.