🎥 AI Video Generation Prompts Pack — 30 Tool-Agnostic Prompts for Higgsfield, Runway, Luma, Kling

Your AI Video Generation Session — Morni Founder Ad (Higgsfield 6s)

Tool Match Check

Higgsfield is the RIGHT tool for this shot. Here's why:

✅ Motion control: Your 'slow dolly in during the action' is exactly what Higgsfield's motion control system specializes in (March 2026 Similarweb data shows them #1 in video-gen traffic largely because of this capability).

✅ Cinematic realism at 6s duration: Higgsfield handles 5-7 second cinematic clips strongly. This is their reliability zone.

✅ Character consistency within single clip: Single-clip character coherence is reliable. Consistency across 4 separate clips (your multi-clip sequence) will need careful prompting — we'll address.

⚠️ Brand palette enforcement: Higgsfield sometimes drifts color palette. We'll lock amber + forest green explicitly.

❌ Specific product (Morni bottle) rendering: Don't rely on Higgsfield to accurately reproduce your specific bottle. This clip keeps product off-screen (smart choice). Future clips featuring bottle should use real product photography composited later.

Shot Design Decisions

Shot type: Medium shot, transitioning to medium close-up via the dolly

Camera movement: Slow dolly-in (35mm equivalent), approximately 18-24mm over the 6 seconds

Lens character: 35mm natural, shallow-to-medium depth of field

Lighting: Golden hour morning light, warm amber cast, soft shadows, window-backlit slightly

Motion pace: Slow and deliberate — pacing reflects 'calm mindful morning' brand feel

Timing breakdown within 6 seconds:

• 0.0-1.5s: Jenny stands, cup beginning to lift
• 1.5-3.5s: Cup rises, meets lips, slow sip
• 3.5-5.0s: Cup lowers slightly, Jenny looks toward window
• 5.0-6.0s: Subtle smile forms, camera is now closer (medium close-up)

The Prompt (paste into Higgsfield)

Cinematic 6-second clip, medium shot transitioning to medium close-up via slow dolly-in.

CHARACTER LOCK: Jenny, woman mid-30s, shoulder-length loose brown hair (slightly messy), round black-framed glasses, navy blue hoodie. Slim build. Consistent appearance throughout the 6-second clip.

SETTING: Minimalist kitchen with warm morning light streaming through window on right side of frame. Wooden countertop visible. Small plant on counter. Subtle amber + forest green accents in background (curtains, ceramic mug, small details).

ACTION BEAT 1 (0-1.5s): Jenny stands by kitchen counter, both hands lifting a ceramic tea cup up toward her face. Slow deliberate motion. Natural, mindful.

ACTION BEAT 2 (1.5-3.5s): Cup reaches her lips. She takes a slow sip, eyes soft, presence felt. Steam rises subtly from cup.

ACTION BEAT 3 (3.5-5s): She lowers the cup slightly, turns her head to look out the window on her right. Morning sunlight catches her face.

ACTION BEAT 4 (5-6s): A subtle satisfied smile forms on her face. Calm and grounded.

CAMERA: Slow continuous dolly-in from medium shot to medium close-up across the full 6 seconds. 35mm equivalent lens. Shallow depth of field with soft background bokeh. Camera movement is smooth, steady, unhurried.

LIGHTING: Golden hour warm morning light from window (right-side key light). Soft shadows. Warm amber cast throughout. Slight window flare acceptable.

STYLE: Cinematic realism. Indie short-film aesthetic — warm, intimate, grounded. NOT corporate, NOT stock-photo-lit, NOT over-produced. Think A24 cinematography meets wellness brand campaign.

COLOR PALETTE: Warm amber tones dominant, subtle forest green accents, natural skin tones. Slightly desaturated but warm. NO oversaturated colors.

MOTION PACE: Slow, deliberate, contemplative. Reflects mindful morning ritual, NOT caffeinated energy.

End frame: Jenny in medium close-up, subtle smile, looking toward window, warm morning light on her face.

Expected Output + Quality Control

Higgsfield will likely produce:

• Strong cinematic motion with smooth dolly-in (Higgsfield's flagship capability)
• Character appearance roughly matches within the 6-second clip (single-clip consistency is reliable)
• Warm amber lighting correctly rendered (explicit specification helps)
• Slight drift on glasses shape or hair may occur — acceptable within clip, verify across subsequent clips

Verify on first generation:

• Jenny's appearance consistent start-to-end of clip
• Camera movement smooth (not jerky or jarring)
• Action timing matches beats (cup lift → sip → window-look → smile flows naturally)
• Background doesn't morph mid-clip (Higgsfield occasionally drifts setting details)
• No text appearing on objects (Higgsfield text rendering is weak — there shouldn't be any text anyway)

What will likely need iteration:

• The specific 'subtle smile' beat — AI often overshoots to 'wide smile' or undershoots to 'flat expression'
• Exact dolly speed — sometimes too fast, iterate with 'extremely slow dolly'
• Hand posture on the cup — AI can produce awkward hand shapes on close inspection

Alternative Tool Recommendations

If Higgsfield motion feels off:

• Runway Gen-3 Alpha — also strong cinematic. Try if Higgsfield output has character-drift issues.

If character consistency is critical across all 4 clips in your sequence:

• Generate all 4 clips in Higgsfield one after another with SAME character lock text in each prompt
• Alternative: Kling AI for longer single clips that maintain character better than multi-generation

If you need more physics realism (e.g., tea actually looking like liquid):

• Luma Dream Machine has stronger physics model. Use for close-up product-motion shots specifically.

If budget is constraint:

• Kling free tier handles simple shots reasonably well
• Pika for stylized/playful mood

Post-Generation Workflow

1. Generate 3-5 variants of this 6-second clip. Character drift across generations means you want options.

2. Cross-reference with other 3 clips in sequence. Your character Jenny needs to look consistent across all 4 scenes. Pick the clip variant that best matches the others.

3. Edit in DaVinci Resolve / Premiere / Final Cut:

- Stitch 4 clips in sequence

- Apply color grade for consistent palette (even if individual clips drift slightly, shared grade unifies)

- Add soft ambient music track (no dialogue needed per brief)

- Add subtle SFX: bird song, distant city ambience, cup click

- Brand logo (Morni) added in post — NOT generated in-clip

4. Audio considerations:

- Higgsfield generates without audio by default

- Background music: Artlist / Musicbed / Epidemic Sound (wellness-appropriate tracks)

- Keep dialogue-free for social media auto-play contexts

5. Export specs:

- Instagram/TikTok: 1080×1920 vertical (9:16), MP4 H.264

- LinkedIn: 1920×1080 horizontal (16:9)

- For Instagram Reels / TikTok: ensure first 1.5s are visually striking (hook)

Safety + Ethics

✅ This is a legitimate use case:

• Marketing your own brand (Morni)
• Character 'Jenny' is fictional (not a real-person likeness)
• Honest brand messaging (calm morning ritual, not deceptive 'before/after' health claims)
• No impersonation of real people

⚠️ FTC considerations:

• Some jurisdictions require disclosure when AI-generated content appears in paid advertising
• EU AI Act (2026 rollout) has specific transparency requirements for AI-generated audio/video
• If you use this in paid ads, check your jurisdiction's current AI disclosure rules

❌ What NOT to do with this pack's techniques:

• Do NOT generate videos depicting real people (even if they 'kind of' look like someone)
• Do NOT generate fake testimonials as if from customers
• Do NOT create 'documentary-style' AI content misleadingly labeled as real events
• Do NOT use for health claims (before/after wellness) that require FDA substantiation

Key Takeaways

• Higgsfield is the right tool for 6-second cinematic clips with motion control — match the tool to the job.
• Motion-beat breakdown (4 beats across 6 seconds) is the pattern that produces coherent action — not a single 'scene description' prompt.
• Character consistency requires explicit lock at start of prompt — for multi-clip sequences, same lock text in every clip.
• Generate 3-5 variants — pick the one that best matches your other clips in the sequence. Never rely on first generation.
• Keep product off-screen in AI clips — composite real product photography in post. AI-generated specific products drift in accuracy.
• Short clips (3-6s) beat long clips (10+s) for quality. Split longer sequences into shorter clips + edit.
• Post-production in Resolve/Premiere handles what AI can't: consistent color grade, music, SFX, brand logo overlay, text elements.
• FTC/EU AI Act disclosure considerations for paid ads — know your jurisdiction's rules before running AI-generated ads.