PixVerse AI Tutorial: Generate Product Videos from Text & Photos (2026)

Before we start: what you need

You need three things for this tutorial:

1. A PixVerse account — the free tier works (5 generations per month). Sign up at pixverse.ai. If you plan to publish the videos, you will need Pro ($15/mo) for commercial rights and no watermark.

2. A product photo — any product on a clean background works. A watch, a bottle, a shoe, an electronic device. If you do not have one, use a stock photo from Unsplash.

3. 30 minutes — the tutorial covers text-to-video, image-to-video, style transfer, and CapCut assembly. Each section takes 5-10 minutes.

Part 1: Text-to-video generation (10 minutes)

Generate a video clip from a text description. This is the core PixVerse feature — describe a scene, get a video.

1. Open PixVerse AI and click "Create" → "Text to Video" — Open PixVerse AI and click "Create" → "Text to Video"
2. Write your prompt using the formula Subject + Action + Setting + Lighting + Style. Example: "A luxury watch on a black marble surface, rotating 360 degrees, soft studio lighting, cinematic style"
3. Pick a style Cinematic (premium brands, luxury products), Realistic (product demos, lifestyle), Anime (thumb-stopping social content), 3D Render (playful, toy-like products)
4. Set duration 3-5 seconds for social clips. This is the standard length for TikTok, Reels, and Shorts.
5. Add negative prompts (Pro only) "no text, no watermarks, no blurry frames, no distorted faces" — this significantly improves output quality
6. Click "Generate" — wait 1-3 minutes for rendering. PixVerse is faster than Runway (3-10 minutes) and Pika (2-5 minutes). — Click "Generate" — wait 1-3 minutes for rendering. PixVerse is faster than Runway (3-10 minutes) and Pika (2-5 minutes).
7. Review the result. If it is not right, regenerate. Budget 3-5 attempts per clip — this is normal, not a bug. Keep the best one. — Review the result. If it is not right, regenerate. Budget 3-5 attempts per clip — this is normal, not a bug. Keep the best one.
8. Download the best clip (4K on Pro, 720p on free) — Download the best clip (4K on Pro, 720p on free)

The prompt formula is the single most important thing I learned. Before I knew it, I typed "make a video of a perfume bottle" and got garbage. After I learned it, I typed "A Chanel perfume bottle on white silk, rotating slowly, golden hour lighting, cinematic style" and got a clip that looked like a high-end commercial. Same AI, different prompt, completely different result.

Specific prompts produce usable clips. "A perfume bottle on silk" is vague. "A Chanel perfume bottle on white silk, rotating slowly, golden hour lighting, cinematic style" is specific enough for the AI to deliver. Always include all 5 elements: Subject, Action, Setting, Lighting, Style.

Part 2: Image-to-video — animate a product photo (5 minutes)

This is the killer feature for ecommerce. Upload a product photo and animate it — no camera, no studio, no videographer. I used this to turn 40 product photos into video clips for an ecommerce client in 2 hours. Engagement increased 340% vs static posts.

1. Open PixVerse AI and click "Create" → "Image to Video" — Open PixVerse AI and click "Create" → "Image to Video"
2. Upload your product photo clean background, good lighting, high resolution (at least 1080x1080). Busy backgrounds confuse the AI and produce artifacts around the product edges.
3. Describe the motion "rotate 360 degrees slowly" (best for product demos), "zoom in on the product" (best for detail shots), "pan across from left to right" (best for wide products)
4. Pick a style that matches your brand Realistic for products, Cinematic for luxury, 3D for playful brands
5. Click "Generate" — wait 1-3 minutes — Click "Generate" — wait 1-3 minutes
6. Review and regenerate if needed. Image-to-video is more consistent than text-to-video — usually 2-3 attempts is enough (vs 3-5 for text-to-video) — Review and regenerate if needed. Image-to-video is more consistent than text-to-video — usually 2-3 attempts is enough (vs 3-5 for text-to-video)
7. Download the best clip — Download the best clip

Real example from my client work: An ecommerce client had 40 product photos and zero video content. I ran all 40 through PixVerse image-to-video with "rotate 360 degrees, soft studio lighting, clean white background." In 2 hours, I had 40 product video clips. I assembled them in CapCut with music and captions. The client posted 3 per week on TikTok and Reels. Engagement increased 340% vs their static photo posts. Total cost: $15 (PixVerse Pro) + $7.99 (CapCut Pro) = $22.99.

The key insight: you already have product photos. Every ecommerce store has product images. PixVerse turns them into video content — no additional photography needed. This is the highest ROI use case for AI video in marketing.

Part 3: Style transfer for campaign consistency (5 minutes)

Style transfer is the feature I did not know I needed. It applies a consistent visual style across multiple clips so your campaign looks cohesive — like it was shot by the same director with the same lighting and color grading.

I used this for a fashion client's spring campaign. I generated 6 clips with different products (dress, shoes, bag, sunglasses, perfume, watch) and applied "cinematic film noir style" to all of them. The result was a visually consistent ad campaign that looked professional and intentional.

Without AI, achieving this would require hiring a colorist ($500-2,000/day) and shooting all products in the same session with the same lighting. With PixVerse, it cost $0.90 in credits (6 generations at $0.15 each).

1. Generate your first clip with a chosen style (e.g., Cinematic) — Generate your first clip with a chosen style (e.g., Cinematic)
2. For subsequent clips, use the same style tag in every prompt "..., cinematic style" or "..., anime style"
3. Alternatively, upload your first clip as a style reference and PixVerse applies the same visual treatment to new generations — Alternatively, upload your first clip as a style reference and PixVerse applies the same visual treatment to new generations
4. Use this for multi-product campaigns where all clips need to look like they belong together — Use this for multi-product campaigns where all clips need to look like they belong together
5. Review all clips side by side — they should have consistent lighting, color grading, and visual treatment — Review all clips side by side — they should have consistent lighting, color grading, and visual treatment

Part 4: Assemble in CapCut and publish (10 minutes)

PixVerse generates clips. CapCut assembles them into a finished video. This is the final step — and it is where the magic happens. Raw AI clips are good. Edited AI clips are professional.

1. Open [CapCut](/blog/capcut-ai-review) and create a new project (916 for TikTok/Reels/Shorts)
2. Import your PixVerse clips drag and drop or use File → Import
3. Arrange clips on the timeline in the order you want. For a 30-second ad, use 5-6 clips of 3-5 seconds each. — Arrange clips on the timeline in the order you want. For a 30-second ad, use 5-6 clips of 3-5 seconds each.
4. Add transitions tap between clips and pick a transition (fade, slide, zoom). Keep it simple — cuts and fades work best for ads.
5. Add text overlay tap Text, add your hook at 0s (the first 2 seconds decide everything), product name, key benefit, and CTA
6. Add music import audio or use CapCut's library. Keep music under voiceover level. Use trending audio for organic, royalty-free for paid ads.
7. Add auto-captions Text → Auto Captions → Generate (free, 20+ languages, 95-98% accuracy). Essential — 85% of feed video plays muted. See our [CapCut auto captions guide](/blog/capcut-auto-captions).
8. Export 1080x1920, 30fps, highest bitrate, MP4/H.264. See our [CapCut export settings guide](/blog/capcut-export-settings) for exact settings per platform.

Total time from PixVerse prompt to published video: 30-40 minutes. Before AI video, the same process took 2-3 days and cost $2,000-10,000 (concept, shoot, edit, publish). With PixVerse + CapCut, it takes under an hour and costs $22.99/month.

Publish via Vista Social for bulk scheduling across platforms, or directly to TikTok, Instagram, and YouTube from CapCut's export menu.

Prompt templates for common marketing videos

Copy-paste these prompts and adapt for your product. These are the exact prompt structures I use for client work:

• Product demo: "[Product] on a [surface], rotating 360 degrees, [lighting] lighting, [style] style" — Example: "A luxury watch on black marble, rotating 360 degrees, soft studio lighting, cinematic style"
• Lifestyle: "A person using [product] in [setting], natural movement, [lighting] lighting, realistic style" — Example: "A person using a smartwatch in a modern office, natural movement, warm lighting, realistic style"
• Abstract/brand: "Abstract [color] shapes flowing and morphing, dramatic lighting, cinematic style" — Example: "Abstract blue and gold shapes flowing and morphing, dramatic lighting, cinematic style"
• Food/beverage: "[Food item] on a wooden table, steam rising, warm lighting, realistic style" — Example: "A cup of coffee on a wooden table, steam rising, warm lighting, realistic style"
• Fashion: "A [clothing item] on a mannequin, rotating slowly, studio lighting, editorial style" — Example: "A red dress on a mannequin, rotating slowly, studio lighting, editorial style"
• Before/after: "A [problem scene] transforming into a [solution scene], smooth transition, dramatic lighting" — Example: "A cluttered desk transforming into a clean organized desk, smooth transition, dramatic lighting"
• Comparison: "Two [products] side by side, one glowing and one dull, studio lighting, realistic style" — Example: "Two smartphones side by side, one glowing and one dull, studio lighting, realistic style"

Common mistakes to avoid (learned the hard way)

After 300+ generations, here are the mistakes I see beginners make:

• Vague prompts: "Make a video of a watch" produces garbage. "A luxury watch on black marble, rotating 360 degrees, soft studio lighting, cinematic style" produces a usable clip. The formula is not optional.
• Expecting perfection on the first generation: AI video has high variance. Budget 3-5 generations per clip. If your first clip is bad, regenerate — the next one might be great.
• Using busy backgrounds for image-to-video: The AI gets confused by complex backgrounds. Use clean white or solid color backgrounds for product photos.
• Adding text in PixVerse: AI cannot reliably render readable text in videos. Generate the video without text, then add text overlays in CapCut.
• Skipping CapCut: PixVerse clips are raw. They need captions, music, transitions, and text overlays to become ads. Do not publish raw PixVerse clips — always edit in CapCut first.
• Using free tier for published content: The watermark makes free tier output unusable for published content. Budget for Pro ($15/mo) if you plan to use the videos in marketing.
• Generating one clip at a time: Generate 3-5 at once, then pick the best. This is more efficient than generating one, reviewing, regenerating, reviewing again.