Image to Video: How to Animate a Product Photo
Biggsfield team · October 18, 2026 · 6 min read
A strong product photo is a finished first frame, so animating it is mostly a matter of deciding where the shot ends, how the camera moves and which model renders it. This guide covers start and end frames, camera motion presets, model choice, prompt structure and ready-to-adapt example prompts in Biggsfield.
Why start from a photo instead of a text prompt
When you generate a product video from text alone, the model invents your product. The label drifts, the cap changes shape and the color shifts. Starting from a real photo fixes the most important variable before anything moves. The packshot you already approved becomes frame one, and the model only has to work out what happens next.
In Biggsfield this happens in the AI Video studio (/studio/video), which lets you animate any photo or lock both the first and the last frame of a shot. Treat your photo the way a cinematographer treats a locked-off opening frame. Composition, lighting and product placement are settled. Your job is to describe motion, camera and sound.
Prepare the photo before you animate it
Motion makes every flaw in a still more obvious, so clean the image first. If the shot was taken on a phone, the Product Photo app turns an everyday product photo into a commercial shot. Object Remover erases stray cables or price tags. Background Swap drops the product into a new location, and Relight changes the light to golden hour, studio or neon before you add any motion.
Leave room in the frame for the move you are planning. A push-in needs a product that is not already filling the edges, and an orbit works best when the product sits near the center with some background around it. If your crop is too tight, Expand Image widens the frame and extends the scene. Image resolution matters too. Nano Banana 2.1 and FLUX 3 both go up to 4K, and the Topaz Upscale app sharpens a still for 14 credits.
Start frame vs. start and end frame
A start frame alone gives the model freedom. It sees your product and invents the rest of the shot from your prompt. Use this for atmospheric motion, such as steam rising from a coffee cup, light sweeping across a watch face or fabric moving in a breeze. The product stays recognizable because the shot begins from your real image.
Locking both the first and last frame gives you control. The model has to travel from image A to image B, so you decide exactly where the shot lands. This is the right choice for reveals: closed box to open box, bottle on its side to bottle upright, wide room shot to tight hero frame. You can build the end frame from the start frame with an image edit, for example with Nano Banana 2.1. Say exactly what changes and what must stay identical, and the product will match between the two frames.
If you want a stylized move between two different images rather than a continuous physical shot, the Transition Video app creates a smooth, creative transition between two images without a prompt.
Choosing the model: Kling 3.0 Pro, Seedance 2.5 or MiniMax H3 Max
Kling 3.0 Pro is the value pick for most product work. It renders 1080p with native audio and costs 140 credits for 5 seconds or 280 credits for 10 seconds (about $1.35 and $2.69 on Creator). Kling responds to physical, concrete action and strong camera language, which suits product shots. Write 'the cap twists off and condensation runs down the glass' rather than 'a refreshing moment'. Its native voice speaks only English or Chinese, so write any dialogue in English. Kling 3.0 also supports multi-shot: give each shot its own self-contained prompt and a duration.
Seedance 2.5 is the flagship. It runs up to 30 seconds, generates native audio with lip-synced speech, sound effects and ambience, and accepts multiple references. In the prompt you refer to reference images as @Image1 and @Image2, which helps when you want the product and a presenter in the same scene. You can write several shots in order as 'Shot 1: … Shot 2: …', each with its own camera. Prices start at 184 credits for 5 seconds at 480p with sound and go to 395 credits at 720p and 970 credits at 1080p. Draft at 480p and render the final at a higher resolution.
MiniMax Hailuo H3 Max is the choice when a person carries the shot: a hand model applying serum, a creator reacting to the first sip. It is strongest at micro-expressions, gestures and natural body weight. It has no native sound, so plan to add audio in the editor. It costs 67 credits for 5 seconds at 768P or 134 credits at 1080P.
If you are unsure which to use, the AI Video studio can run one idea on up to four models at once so you can keep the best take. For a product that has to stay identical across several clips, Kling O3 Pro is built for character and product reference video and keeps up to four references consistent.
Camera motion presets
The AI Video studio includes more than 20 camera moves, among them crash zooms, dolly zooms, FPV and bullet time. Each one is added to the prompt in that model's own wording. This matters because models respond to different phrasing, and the built-in AI director writes the version each model understands.
Match the move to the message. A slow push-in signals premium and focus, which suits cosmetics, watches and fragrance. An orbit shows form and is good for sneakers, bottles and devices. A crash zoom adds energy for food, drinks and anything aimed at short-form feeds. FPV moves suit spaces such as hotel rooms, interiors and stores. Use one move per shot. Stacking an orbit, a tilt and a zoom in five seconds usually produces a warped product.
Prompt tips that keep the product intact
Do not re-describe the photo in detail. The model can already see it. Spend your words on what changes: the action, the camera, the light over time and the sound.
State what must not change. A line such as 'the label, logo and bottle shape stay exactly as in the image' is worth including in every product prompt.
Write motion physically. Liquids pour, foil crinkles, light glints across brushed metal, and dust drifts through a backlight. Concrete verbs give better physics than adjectives.
Direct the audio. With Kling 3.0 Pro or Seedance 2.5, describe the sound: the snap of a lid, fizz, footsteps on concrete. For Seedance dialogue, put the line in double quotes and say who speaks it and how.
Keep durations honest. Five seconds fits one action and one camera move. If you need a longer sequence, use multi-shot prompts, or extend a clip from its last frame with Extend Video on MiniMax H3 (up to 15 seconds) or FLUX 3 (up to 20 seconds with audio).
Example prompts
Adapt these to your own product. The first is written for Kling 3.0 Pro, the second for Seedance 2.5 with a reference image, and the third for MiniMax H3 Max.
Slow push-in on the perfume bottle from the image. A beam of warm side light sweeps left to right across the glass, catching the facets. Fine mist drifts behind the bottle. The label and bottle shape stay exactly as in the image. Sound: a soft airy hum and a faint glass chime.
Shot 1: Close-up of @Image1, a can of sparkling water on a wet stone ledge. The tab cracks open and bubbles rise, with a slow crash zoom toward the opening. Shot 2: Medium shot, a woman lifts the can and says, smiling, "This is my afternoon reset." Sound: crisp can opening, fizz, distant birdsong.
A woman holds the serum bottle from the image at eye level, studies it, then breaks into a small satisfied smile. She tilts her head, presses the dropper and lets one drop fall onto her fingertip. Her shoulders relax as she exhales. Camera: gentle handheld push-in from a medium shot to a close-up of her face and the bottle.
Orbit around the white sneaker from the image, placed on a concrete plinth. Hard late-afternoon sunlight casts a long shadow that rotates with the camera. Laces shift slightly in a breeze. The logo and stitching stay identical to the image. Sound: wind and distant city traffic.
Shortcuts and finishing
For common product formats you may not need a prompt at all. The 360° Product Video app creates a cinematic spin, Splash Packshot puts the product in a slow-motion splash of water, juice or paint, and Unboxing Video shows hands unboxing the product on camera, ASMR style. For directional control, Biggs Sketch lets you draw arrows on the photo. An arrow on the subject sets its path, and an arrow from the edge of the frame sets the camera move.
Finish in Biggs Cut. Trim takes, add music and voiceover, burn in word-timed subtitles and export 9:16, 1:1 and 16:9 versions from the same project. If a clip needs more resolution, Topaz Video Upscale takes it to 4K for 134 credits per 5 seconds.
Questions
Which models can I use to animate a product photo?
The AI Video studio lets you animate any photo or lock both the first and last frame of a shot. Seedance 2.5 accepts multiple reference images that you call @Image1, @Image2 in the prompt. Kling O3 Pro keeps products consistent across up to four references. Kling 3.0 Pro and MiniMax H3 Max are strong options for physical product motion and human performance.
How much does a 5-second animated product clip cost?
On Kling 3.0 Pro it costs 140 credits with sound. MiniMax H3 Max costs 67 credits at 768P or 134 credits at 1080P, without sound. Seedance 2.5 ranges from 184 credits at 480p to 970 credits at 1080p, both with sound.
How do I stop the product from warping or the label changing?
Start from a clean, high-resolution photo and use one camera move per shot. Describe physical action, not mood, and state in the prompt that the label, logo and shape stay exactly as in the image. For reveals, lock an end frame so the model has to land on your approved product.