Restaurant Taste AI Video Generator
Generate a restaurant-table video of a woman lifting pasta to her mouth at an outdoor cafe, shot like a still phone clip.
What this Restaurant Taste template creates
What you create
Generate a static-camera outdoor restaurant clip of a pasta bite
Defining visual direction
An outdoor cafe table, pasta, orange drinks, and a still-phone camera define the shot.
Best for
Food-lifestyle clips, cafe social videos, and UGC-style ads
What to upload for Restaurant Taste
Prepare a text prompt
No image is required. Describe one diner, one table setting, and one slow bite.
Keep the core composition
An outdoor cafe table, pasta, orange drinks, and a still-phone camera define the shot.
Use the tested starting point
Generate the baseline first, then change only one creative variable.
MiniMax H3 prompt and settings
Model
MiniMax H3 (minimax_h3)
Verified settings
7s, 720p, 9:16, text only, audio off
Reusable English prompt
Create one continuous video. Vertical 9:16 aspect ratio video. A highly photorealistic, casual smartphone-style video with a completely static camera. A glamorous young woman [@reference] with sleek dark hair in a low bun, black narrow sunglasses, and gold hoop earrings sits at an outdoor restaurant table. She is wearing a stylish, form-fitting black top with an elegant neckline. On the table is a plate of pasta and orange drinks with black straws. She gracefully brings a fork with long noodles to her mouth, gently taking a bite and chewing with slow, minimal, elegant movements. Her body maintains subtle, natural breathing motion. She maintains a calm expression, looking slightly to the side. The background is a bright luxury shopping arcade with a glass ceiling, ornate beige architecture, and softly blurred pedestrians. Natural daylight, realistic shadows, warm highlights, slight vintage film grain, pristine live-action realism strictly avoiding any glossy CGI or artificial plastic aesthetics. @MyCharacter5
How to create Restaurant Taste
Generate a static-camera outdoor restaurant clip of a pasta bite Follow the three-step text to video workflow and review the first video before iterating.
1. Add a text prompt
No image is required. Describe one diner, one table setting, and one slow bite.
2. Generate the baseline
Keep the defining direction: An outdoor cafe table, pasta, orange drinks, and a still-phone camera define the shot.
3. Review and refine
Check hands, fork, noodles, glasses, and face continuity through the bite. Then change the dish or outfit, but keep the static camera and outdoor table.
Before you publish Restaurant Taste
Check hands, fork, noodles, glasses, and face continuity through the bite. Results can vary between generations, so inspect the full-size video before publishing.
Match the intended result
The video should generate a static-camera outdoor restaurant clip of a pasta bite.
Inspect the details that matter
Check hands, fork, noodles, glasses, and face continuity through the bite.
Change one variable
Change the dish or outfit, but keep the static camera and outdoor table.
Where Restaurant Taste works well
Useful for food-lifestyle clips, cafe social videos, and UGC-style ads.
Cafe Lifestyle Clips
Keep the camera still so it feels like a phone clip.
Food Social Videos
Lead with the pasta bite as the action.
UGC-Style Ads
Review glasses and hands before posting.
Restaurant Taste questions
Answers about inputs, settings, customization, API reuse, and output formats for Restaurant Taste.
/ 01What does this template create?
Generate a static-camera outdoor restaurant clip of a pasta bite
/ 02What input is required?
No image is required. Describe one diner, one table setting, and one slow bite.
/ 03Which model and settings are used?
MiniMax H3 with 7s, 720p, 9:16, text only, audio off.
/ 04How can I customize it?
Change the dish or outfit, but keep the static camera and outdoor table.
/ 05Can I reuse this template through the API?
Yes. Send model key minimax_h3 and the reusable English prompt from this page to create programmatic variations.
/ 06Which output format should I expect?
Image templates return a generated image such as PNG or JPEG; video templates return an MP4-style video. Review the final output and resize only after generation.
/ 07Will every result match the demo?
No. Check hands, fork, noodles, glasses, and face continuity through the bite. Details vary between generations.