Prompts
Ten copy-ready social media AI video prompts on muvi.video for 9:16, 1:1, and 16:9, with per-platform aspect-ratio guidance for TikTok, Reels, Shorts, and feed.
To validate prompt adherence, hook pacing, and human kinematics on mobile feeds, here is a verified generation from Muvi's benchmark suite:
“9:16 vertical full-bleed commercial, rapid dynamic camera push-in on a creator unboxing innovative tech gear, expressive eye contact, dramatic modern studio edge lighting, 24fps motion blur, 1080p Full HD”
Notice the rapid hook delivery in the first two seconds, zero anatomy distortion during quick hand movements, and pristine background bokeh.
A social video does not fail because the idea was weak. It fails because the format was wrong: a 16:9 clip cropped to vertical loses its subject, a 9:16 clip stretched to feed leaves dead bars, and a slow opening loses the viewer before the hook lands. Getting a social result from an AI video generator starts with one decision made before anything else: the aspect ratio, chosen per platform.
On muvi.video, you are not locked into a single proprietary model. You have an entire studio of 20+ top-tier engines:
This page explains social prompt anatomy, platform-specific aspect ratios, model decision logic, and ten copy-ready production prompts spanning all three formats.
A social AI video prompt is not a one-line caption, it is a layered specification tuned for the feed. The components below differ from cinematic prompting on purpose, social rewards a fast hook, a readable subject at thumb-scroll speed, and an aspect ratio chosen for where the clip will play. Generic style adjectives add noise, explicit instruction guides output.
1. Aspect Ratio and Platform Format This is the first-sentence decision, not an afterthought. State the ratio before anything else, "9:16 vertical, full-bleed framing" for TikTok, Reels, and Shorts, "1:1 square, centered subject" for feed, or "16:9 landscape" for in-feed and wide placements. Veo 3.1 outputs 16:9 and 9:16. Seedance 2.0 adds 1:1, 4:3, 3:4, and 21:9, so route any square or non-Veo ratio to Seedance. Choosing the ratio first forces every later instruction to compose for the correct frame.
2. The Hook (First 1 to 2 Seconds) Social feeds are won or lost in the opening beat. Describe what the viewer sees in the first second, "opens on an extreme close-up of hands cracking an egg into a hot pan, immediate sizzle" gives the model a strong first frame. A clip that opens on a slow establishing wide loses the scroll. Put motion or a visual payoff at the very start.
3. Subject Scale and Safe Zones Vertical and square frames crop tighter than landscape, and platform UI overlays the edges, captions sit at the bottom, the profile and action buttons sit on the right of 9:16. Instruct the model to keep the subject "centered with headroom above, action contained in the middle third, lower third clear for captions." This keeps the subject readable on a phone and out from under the interface.
4. On-Screen Motion and Pacing Social pacing is faster than cinematic. Specify a single clear motion arc that resolves inside the clip length, "quick push-in over 3 seconds, ending on the product label" or "subject turns to camera and smiles within the first 2 seconds." Avoid multi-beat camera choreography that a short clip cannot finish, it interpolates into something that looks unfinished.
5. Native Audio Strategy Veo 3.1 generates native ambient and dialogue audio alongside the video. For social, decide upfront whether the clip carries its own diegetic sound, "ambient kitchen sizzle and chopping, no music," or whether it will sit silent under a trending track added in the editor. State "no spoken dialogue, ambient sound only" when you plan to drop a music bed over the clip in post, so the model does not generate competing speech.
6. Color and Brand Read Social color needs to read at small size and in bright daylight on a phone. Name a high-contrast, saturated-but-clean palette, "bright daylight, punchy saturated colors, clean white background" reads instantly in feed. Muted prestige grades that look great on a cinema screen disappear at thumbnail scale. If the clip carries a brand, name the dominant brand color as an accent in frame.
Aspect ratio is the single highest-leverage social decision, and it is platform-specific. The guide below maps each major surface to its native ratio and the model that outputs it on muvi.video. Generate at the target ratio, do not crop a 16:9 clip down to vertical, cropping reframes the composition and discards resolution the original prompt intended.
9:16 vertical, TikTok, Instagram Reels, YouTube Shorts The dominant short-form format. Full-screen vertical, subject centered, lower third kept clear for captions and the platform caption bar. Veo 3.1 outputs 9:16 natively at 1080p with audio, the default choice for vertical short-form. Seedance 2.0 also outputs 9:16 when you need a clip longer than 8 seconds.
1:1 square, Instagram and Facebook feed The square feed post still earns more vertical real estate in a scrolling feed than 16:9 while staying balanced top and bottom. Veo 3.1 does not output 1:1, so route every square clip to Seedance 2.0, which supports 1:1 across durations from 4 to 15 seconds. Compose with the subject centered, since a square frame has no strong directional bias.
16:9 landscape, in-feed video, YouTube standard, X and LinkedIn Landscape still wins for in-feed players, YouTube standard placement, and professional networks where wide framing reads as polished. Veo 3.1 outputs 16:9 natively at 1080p with audio. Use it for talking-head clips, product demos that need width, and any cross-post that will also live on YouTube.
4:3 and 3:4, niche and legacy placements Some feeds and ad placements still accept 4:3 or its vertical sibling 3:4. Only Seedance 2.0 outputs these on muvi.video. Reach for them only when a specific placement requires it, otherwise default to 9:16, 1:1, or 16:9.
21:9 ultrawide, cinematic social and banners For a cinematic letterbox look in a feed, or a wide banner clip, Seedance 2.0 outputs 21:9. Treat it as a stylistic choice, not a default, most feeds will pillarbox or crop it.
Not every social clip needs the same model. The tree below maps the most common social cases to the model that handles them best. Choosing the wrong model costs coins and produces output that prompt revision alone cannot fix, the most common error is asking Veo 3.1 for a 1:1 square it does not output.
Veo 3.1, Best for: 9:16 and 16:9 social at 1080p with native audio
Veo 3.1 generates clips of 4, 6, or 8 seconds at 1080p (1920x1080) in 16:9 or 9:16, with native ambient and dialogue audio, and returns 4 variants per run so you can pick the best take without re-spending. It does not output 1:1, 4:3, 3:4, or 21:9, and it does not support deterministic seed control, each run with the same prompt may vary. For multi-clip consistency, keep character, wardrobe, and setting descriptions identical across prompts, or use the Veo 3.1 Image-to-Video variant to anchor frames with a reference still. Standard and Fast variants are available, Standard for final delivery, Fast for cheap iteration. For current pricing and coin costs, see the pricing page. For full capabilities, see the Veo 3.1 model page.
Use Veo 3.1 when: you need 9:16 vertical for TikTok, Reels, or Shorts, or 16:9 landscape, at 1080p, your clip is 8 seconds or shorter, and you want native audio.
Seedance 2.0, Best for: square feed, longer clips, and every non-Veo aspect ratio
Seedance 2.0 generates clips at any integer duration from 4 to 15 seconds, the widest duration range in the catalog, and it supports six aspect ratios (1:1, 9:16, 16:9, 4:3, 3:4, 21:9), the broadest format range on muvi.video. It outputs at 480p or 720p, with the Start/End Frame variant up to 1440p. It also offers omni-reference for carrying a subject or style across generations. This is the model for square 1:1 feed posts, since Veo 3.1 does not output 1:1, and for any social clip that needs to run longer than 8 seconds in a single take.
Use Seedance 2.0 when: you need a 1:1 square feed post, a clip longer than 8 seconds, or any of the 4:3, 3:4, or 21:9 ratios. It is the multi-format workhorse for social.
Kling 3.0, Best for: specific stylized motion aesthetics
Kling 3.0 has three variants in the catalog (Text-to-Video, Image-to-Video, Kling O3) and returns 3 variants per run. It handles certain motion-smoothness and stylized aesthetics that render differently on other models. If your reference look is tied to a specific motion texture rather than a hard spec, test a clip on Kling before committing a full set.
Decision summary:
Each prompt below names its aspect ratio and a recommended model. The first frame is written as a hook on purpose, keep that opening beat intact. Swap product, subject, or location to fit your brand, but keep the structural vocabulary, it carries the format and pacing instruction.
Prompt 1, 9:16 Recipe Hook for TikTok and Reels Recommended model: Veo 3.1 Standard (9:16, 1080p, native audio)
9:16 vertical, full-bleed framing. Opens on an extreme close-up of hands cracking an egg into a hot cast-iron pan, immediate sizzle. Quick pull-back over 2 seconds to a medium shot of a bright modern kitchen. Subject centered, lower third kept clear for captions. Bright daylight, punchy saturated colors, clean white counter. Ambient kitchen sizzle and chopping, no music, no spoken dialogue. 1080p.
Prompt 2, 9:16 Product Unboxing for Shorts Recommended model: Veo 3.1 Standard (9:16, 1080p, native audio)
9:16 vertical, subject centered with headroom above. Opens on hands lifting a sleek phone box lid, soft thud of the lid releasing. Slow tilt down revealing the device on a clean pastel background over 4 seconds. Action contained in the middle third, lower third clear for captions. Bright even studio light, saturated accent color on the packaging. Ambient room tone and unboxing foley, no music. 1080p.
Prompt 3, 1:1 Square Feed Product Loop Recommended model: Seedance 2.0 (1:1 square, 4 to 6s)
1:1 square, centered composition. A glass of iced coffee on a marble counter, condensation beading on the glass. Slow 360-degree turntable rotation of the glass over 5 seconds, ice cubes catching the light. No directional bias, subject perfectly centered. Bright daylight, high saturation, clean white-and-marble palette. Smooth seamless motion suitable for a looping feed post. No audio focus, ambient cafe murmur only.
Prompt 4, 1:1 Square Talking Subject for Feed Recommended model: Seedance 2.0 (1:1 square, 6 to 8s)
1:1 square, subject centered with even headroom top and bottom. A friendly presenter in their late 20s faces the camera in a bright home office, gestures naturally as if mid-explanation. Subject framed chest-up, centered, plenty of margin on all sides for a square crop. Soft window light from the left, warm clean color, neutral background. Steady locked-off camera, subtle natural movement only. Suitable for feed autoplay without sound.
Prompt 5, 16:9 Landscape Talking-Head for YouTube and LinkedIn Recommended model: Veo 3.1 Standard (16:9, 1080p, native audio)
16:9 landscape, subject positioned on the left third, clean negative space on the right for on-screen text. A professional speaks directly to camera in a modern office, calm confident delivery for a few seconds. Soft key light from camera left, fill from right, neutral desaturated office background. Shallow depth of field separating subject from background. Ambient room tone, clear spoken delivery, no music. 1080p.
Prompt 6, 9:16 Before-and-After Transition Recommended model: Veo 3.1 Standard (9:16, 1080p)
9:16 vertical, full-bleed framing. Opens on a dull cluttered desk, then a quick whip-pan transition at the 2-second mark reveals the same desk clean and styled. Subject centered, lower third clear for captions. Bright daylight, the messy half slightly desaturated, the clean half punchy and saturated for contrast. Snappy single transition, no slow build. Ambient room tone, no spoken dialogue, leaves space for a music bed in post. 1080p.
Prompt 7, 16:9 Product Demo with Width Recommended model: Veo 3.1 Standard (16:9, 1080p, native audio)
16:9 landscape. Opens on a wide shot of a robot vacuum starting from its dock, immediate motor hum. Slow dolly follows it gliding across a hardwood floor over 6 seconds, gentle low camera angle. Subject leads into negative space on the right, leaving room for a text callout. Bright clean home interior, saturated accent on the device. Ambient floor and motor sound, no music. 1080p.
Prompt 8, 9:16 Fashion Outfit Reveal for TikTok Recommended model: Veo 3.1 Standard (9:16, 1080p)
9:16 vertical, full-bleed framing. Opens on a close-up of shoes stepping into frame, then a quick tilt up over 2 seconds revealing a full outfit, subject mid-stride. Subject centered, action in the middle third, lower third clear for captions. Bright daylight on a clean city sidewalk, punchy saturated wardrobe colors. Single confident walk-and-turn motion. Ambient street tone, no spoken dialogue, space for a trending track in post. 1080p.
Prompt 9, 1:1 Square Animated Quote Card Motion Recommended model: Seedance 2.0 (1:1 square, 4 to 6s)
1:1 square, centered composition. An abstract gradient background of soft orange and deep navy slowly shifting and flowing, gentle organic motion. No human subject, pure motion-graphic texture suitable behind overlaid text. Centered, balanced, no directional pull. Saturated clean palette that reads at small size in feed. Seamless slow loop over 5 seconds. Soft ambient pad tone only, intended to sit silent under added captions.
Prompt 10, 21:9 Cinematic Social Banner Recommended model: Seedance 2.0 (21:9 ultrawide, 6 to 8s)
21:9 ultrawide letterbox. A wide aerial-style glide over a coastline at golden hour, gentle forward push over 7 seconds. Subject is the landscape itself, horizon kept level on the upper third. Warm saturated golden-hour color, clean and bright enough to read at banner size. Smooth single forward motion, no abrupt cuts. Ambient ocean wind, no music, intended as a wide hero banner clip behind overlaid text.
A social clip rarely ships from the first generation. The loop below cuts coin spend by separating a cheap composition pass from the final style-and-format pass.
Use Veo 3.1 Fast to test the hook, subject placement, and the opening beat at your target aspect ratio. Skip expensive style detail here, confirm that the first second reads and that the subject sits clear of caption and UI safe zones. Fast costs fewer coins than Standard.
Once the hook and framing work, switch to Veo 3.1 Standard and add the full social vocabulary, saturated color read, ambient audio strategy, and the safe-zone framing. Because Veo 3.1 returns 4 variants per run and does not support deterministic seed control, generate a run, pick the strongest of the 4 takes, and keep prompt language consistent across clips for a multi-part series.
For a 1:1 square feed post, a 4:3, 3:4, or 21:9 placement, or any clip longer than 8 seconds, move to Seedance 2.0 after confirming the idea in a Veo 3.1 draft where possible. Seedance supports any integer duration from 4 to 15 seconds across all six ratios, with omni-reference to carry a subject across clips. Resolution caps at 720p, or 1440p on the Start/End Frame variant.
Final delivery uses Standard variants generated at the native target ratio. Do not crop a 16:9 output to 9:16 or 1:1, cropping discards resolution and reframes the composition. For a vertical series, render every clip at native 9:16, for a square series, render every clip at native 1:1 on Seedance 2.0.
Consistency across a series: Veo 3.1 does NOT support deterministic seed control, each run with the same prompt may vary. For a multi-clip social series, keep character, wardrobe, setting, and color language identical across prompts, anchor frames with the Veo 3.1 Image-to-Video variant, or use Seedance 2.0 omni-reference to carry a subject across clips.
Use 9:16 vertical for all three. They are full-screen vertical short-form surfaces, so a 9:16 clip fills the frame with no bars and gives the subject the most screen. On muvi.video, generate 9:16 natively on Veo 3.1 at 1080p with audio, or on Seedance 2.0 when you need a clip longer than 8 seconds. Keep the lower third clear for captions and the platform caption bar, and keep the subject out from under the right-side action buttons.
Yes, but use Seedance 2.0, not Veo 3.1. Veo 3.1 outputs only 16:9 and 9:16, it does not output 1:1. Seedance 2.0 supports six aspect ratios including 1:1, at any duration from 4 to 15 seconds. Compose with the subject centered, since a square frame has no strong directional bias, and keep margins even on all sides.
Seedance 2.0 is the multi-format workhorse because it supports six aspect ratios (1:1, 9:16, 16:9, 4:3, 3:4, 21:9), which covers every common social placement, including the 1:1 square that Veo 3.1 cannot produce. When you specifically need 9:16 or 16:9 at 1080p with native audio, Veo 3.1 is the higher-resolution choice. A common workflow is to draft on Veo 3.1 for the 9:16 and 16:9 cuts and use Seedance 2.0 for the square and ultrawide cuts of the same campaign.
Veo 3.1 generates clips of 4, 6, or 8 seconds at 1080p, which fits most short-form hooks and product beats. For a single continuous clip longer than 8 seconds, Seedance 2.0 supports any integer duration from 4 to 15 seconds at a lower resolution ceiling (720p, or 1440p on the Start/End Frame variant). For a longer social piece, you can also cut several 8-second Veo 3.1 clips together in the editor to keep every frame at full resolution.
Put motion or a visual payoff in the first one to two seconds. Describe the opening frame explicitly in your prompt, an extreme close-up, an immediate action, or a fast reveal, rather than a slow establishing wide. A clip that opens on a static wide loses the scroll before the idea lands. The prompts on this page are written with the hook in the first sentence on purpose.
It depends on your edit. Veo 3.1 generates native ambient and dialogue audio. If you plan to add a trending track in the editor, write "no spoken dialogue, ambient sound only" so the model does not generate competing speech under your music bed. If the clip carries its own diegetic sound as the point, like a sizzle or an unboxing foley, let the ambient audio carry it and skip the added track.
Veo 3.1 generations are coin-based across plans, and the included monthly coin allotment differs by tier. The Ultra Yearly plan includes unlimited Veo 3.1, with no watermark on paid plans. For current per-generation coin costs and monthly coin allotments by plan, see the pricing page.
Apply these prompts in muvi.video Studio. Choose Google Veo 3.1 for 9:16 vertical and 16:9 landscape with native audio, ByteDance Seedance 2.5 for 1:1 square feed posts and 15-second continuous takes, and Alibaba Wan 3.0 Prime for crisp typography and 4K Ultra HD exports. Pick the ratio per platform, paste a prompt, and move from first draft to publish-ready social video in one workspace.
No credit card required · 50 Free Welcome Coins · 480p to 4K Ultra HD · Seedance 2.5, Veo 3.1, Wan 3.0 Prime, and Kling 3.0 available
Ten copy-ready social media AI video prompts on muvi.video for 9:16, 1:1, and 16:9, with per-platform aspect-ratio guidance for TikTok, Reels, Shorts, and feed.
