How to Make a Tech Product Video with AI
4-step AI workflow for tech product videos in 2026: Nano Banana Pro for stills, Seedance 2.0 for motion, Veo 3.1 for lifestyle shots. $11 for a full earbud reveal trailer.
You can make a tech product video with AI in 2026 using three models in sequence: Nano Banana Pro for hero stills, Seedance 2.0 for motion and rotation, and Veo 3.1 for lifestyle footage. The full workflow runs under an hour and costs $8 to $14 in model credits. Everything below is the detail.
TL;DR
- Hero stills: Nano Banana Pro from your product photo, studio-lit at 2K, 15 to 20 seconds per image
- Motion and rotation: Seedance 2.0 with multi-reference conditioning, 60 to 90 seconds per clip, $0.50 to $0.75 each
- Lifestyle in-use footage: Veo 3.1 at up to 4K, 8s clips, cinematic look without a film crew
- Full $799 wireless earbud reveal trailer: 9 clips, $11.20 in model credits, produced in 52 minutes
Tech product video categories
Five formats, each with different generation priorities:
Consumer hardware reveal (earbuds, wearables, audio): surface quality is everything. Color accuracy, material texture, and scale relative to a hand or ear. Lives on YouTube pre-roll and product pages.
App launch: AI generation handles the lifestyle context (hands, desk, phone) while screen content is designed and composited in separately.
Accessory ad (cables, cases, docks): high detail requirements on ports and connection geometry. The bottleneck is photorealism on small hardware details.
Smart home device (hubs, cameras, sensors): needs in-situ shots with Veo 3.1's environmental staging. Device in a kitchen or living room, not on a white surface.
AI tool launch: mostly Veo for "flow state" lifestyle shots. No physical product to animate.
4-step workflow
Step 1: Hero product still with Nano Banana Pro
Every tech product video starts with a clean hero image. You need this before you can generate motion. Upload your actual product photo as a reference, then prompt for the shot type you want.
Prompt that produced a clean result for a matte-black earbud case:
Matte black wireless earbud case on a dark grey brushed aluminum surface. Studio three-point lighting, key light from upper-left, rim light catching the hinge. Case lid open, earbuds visible inside. Straight-on angle, 10-degree tilt. No background texture. No text. Ultra-sharp product shot, 2K.
Generated in 18 seconds. The matte finish rendered accurately, the hinge was clean, and the earbuds inside held correct geometry. The model reproduced the reference device's form without adding or removing ports, which matters when port placement is a recognizable product identifier.
Generate 3 to 5 stills with slightly different angles. Pick the best two as motion input for Step 2.
Step 2: Motion and rotation with Seedance 2.0
Pass your best still from Step 1 into Seedance 2.0 as a reference image. Multi-reference conditioning means the model animates your actual product rather than generating a generic substitute.
Prompt for a 180-degree product rotation:
Wireless earbud case rotates slowly 180 degrees clockwise on a dark surface. Studio lighting stays consistent through rotation. Case stays centered in frame. Smooth mechanical rotation, no bounce. 5 seconds. Vertical 9:16.
72 seconds to generate, $0.62. The rotation was smooth and lighting stayed consistent through the full 180 degrees, which is the hard part for AI video. One of 3 variants had a brief artifact at the 90-degree mark, a common failure point; the other two were clean.
Prompt for a lid-open reveal:
Matte black earbud case sitting closed on a surface. The lid opens slowly to reveal the two earbuds inside, backlit slightly. Smooth mechanical motion, 3 seconds. Product stays center-frame. No camera movement.
Generated in 68 seconds, $0.58. The motion was correct on the first try. Use this as the reveal beat in your timeline.
Step 3: Lifestyle in-use footage with Veo 3.1
Veo 3.1 at 4K handles lifestyle shots well: someone putting earbuds in, hands adjusting a smart speaker, a phone on a desk. Environmental staging is strong enough to place your product in a believable real-world context.
Prompt for a putting-in-earbuds shot:
Close-up of a person's ear and jaw, hands lift a small matte black earbud from a case and place it gently in the ear. Morning light from a window, soft and warm. Slight depth of field. Realistic skin, natural motion. 6 seconds. No face shown above the jawline. Vertical 9:16.
Generated in about 90 seconds at 4K. Hands moved naturally, earbud scale relative to fingers was accurate, depth of field looked like a professional shoot. "No face above jawline" kept the shot generalized without triggering Veo's face-accuracy variance.
For smart home and on-desk products, swap the lifestyle context:
A small white hub sits on a kitchen counter near a window. A hand reaches in from frame-left and presses a soft-touch button. LED ring on the device pulses gently. Natural afternoon light. 5 seconds.
Kitchen context, ambient light, and LED behavior all rendered correctly in a single generation. Environmental placement is where Veo 3.1 separates from other models.
Step 4: Captions and export
Add captions at any spoken segment. White text, thin black outline, lower third. 60% to 85% of social plays are sound-off. Rotation and product reveal clips are typically music-only; captions come in at the VO or spokesperson cut.
Export at your Veo 3.1 clip resolution (up to 4K) to preserve quality for platform upscaling.
Routing by product type and audience
| Audience | Use | Skip |
|---|---|---|
| Consumer hardware | All three models | Nothing |
| Prosumer hardware | Nano Banana + Seedance; Veo for environment | Casual lifestyle; use "studio, clinical, precise" |
| Enterprise hardware | Nano Banana + Seedance only | Veo (unless you need an installation scene) |
| App or SaaS launch | Veo for lifestyle; Nano Banana for device mockup | Seedance unless a physical product is in frame |
Consumer needs in-use footage. Prosumer needs port and button detail. Enterprise needs spec visualization. App launches are mostly Veo with screen content composited in.
Walkthrough: $799 earbud reveal trailer for $11.20
One product photo (the actual earbud case, front-facing, neutral background) is the only input. Here's the full generation run:
| Clip | Model | Prompt type | Cost | Time |
|---|---|---|---|---|
| Hero still, front angle | Nano Banana Pro | Studio product | $0.08 | 18s |
| Hero still, 45-degree angle | Nano Banana Pro | Studio product | $0.08 | 17s |
| Rotation 180 degrees | Seedance 2.0 | Multi-ref motion | $0.62 | 72s |
| Lid-open reveal | Seedance 2.0 | Multi-ref motion | $0.58 | 68s |
| Close-up earbud detail | Seedance 2.0 | Static detail shot | $0.61 | 65s |
| Hand + ear, putting in | Veo 3.1 | Lifestyle | $2.40 | 92s |
| Walking exterior, earbuds in | Veo 3.1 | Lifestyle | $2.40 | 88s |
| Morning desk, earbuds out of case | Veo 3.1 | Lifestyle | $2.40 | 91s |
| Product + packaging end card | Nano Banana Pro | Studio product | $0.08 | 19s |
Total: $9 clips, $11.25 in model credits, 52 minutes start to finish.
Assembly took 18 minutes in 8frame Studio using the product reveal template. Final cut: rotation, lid-reveal, detail close-up, three lifestyle cuts, end card. 28 seconds. No footage shot. No studio booked.
Pitfalls
Button and port detail accuracy. Both Seedance and Nano Banana preserve the overall form of your reference device without hallucinating ports. They will soften sharp edge details at small scales. If port placement is a recognizable product feature, zoom your reference crop to the detail area and generate a dedicated tight shot, then cut it in separately.
Logo and text on device. AI models handle logos badly. The wordmark will blur, distort, or be replaced with something plausible but wrong. Fix: add "No text, no branding on device" to the prompt and composite the correct logo in post. Takes about 3 minutes in any NLE.
Glossy black surface reflections. Glossy black is the hardest material for product generation. The reflection geometry will look plausible but geometrically off, which technical audiences notice. Options: use a real photo of the device as your reference input so you're animating a real photo, or use matte and anodized finishes for generated elements and comp the real device at key frames.
Scale relative to hand. In lifestyle clips, the product scale can drift. An earbud case can end up looking like a deck of cards if you don't anchor the prompt. Add explicit size context: "small rectangular case, about the size of a AA battery stood upright" or "earbud the size of a fingernail." If it looks like a toy fair prop, the scale is wrong.
FAQ
What's the photo-realism limit for AI tech product video in 2026?
Matte and satin finishes: close enough for most social and web placements. Glossy black and mirror-finish aluminum: not quite. Reflection geometry reads as off to technical audiences on rotation clips. If your product is highly specular, generate motion clips with matte-finish language in the prompt, then comp your real hero image at key frames.
Can I generate a video of my actual device?
Yes, with a reference photo. Seedance 2.0 and Nano Banana Pro both accept image references and model output on your device's form factor. They reproduce shape, color, material, and general layout. They won't reproduce exact engravings, tight-tolerance port geometry, or trademarked logos. For any element that requires exact fidelity, generate the surrounding context with AI and composite real device assets at the frame level.
Which model is better for an app launch vs a hardware reveal?
App launches: Veo 3.1 for lifestyle context (someone at a desk, phone in hand). Skip Seedance and Nano Banana unless a physical device is in frame. Hardware reveals: all three in sequence, Nano Banana to Seedance to Veo. If your app has a companion physical device, run the full hardware workflow for that element and Veo-only for the digital context shots.
Clone the tech product video template from 8frame's workflow library, upload your product photo, and run Step 1 first. The Nano Banana still takes 20 seconds and tells you whether your reference image is clean enough to work from. That's your gate check before spending on motion clips.
For model routing across other product video formats, see the how to make a Shopify product video with AI guide.