60 original video directing briefs across 10 production categories. Plan text-to-video scenes, animate images, combine references, edit clips conversationally, interpolate endpoints, and extend selected footage with clear continuity and audio instructions.
English · 简体中文 · 繁體中文 · 日本語 · 한국어 · Español · Français · Deutsch · Português do Brasil · Italiano · العربية · Русский · Bahasa Indonesia · ไทย · Tiếng Việt
Static concept artwork created for this collection. It is not a Gemini Omni video output. See the generation record.
- Choose a brief from the complete prompt index.
- Prepare every input named by the brief and assign one clear role to each image or video.
- Replace bracketed variables, then copy the complete prompt into a surface that supports the required workflow.
- Review motion, identity, geometry, timing, audio, text, and the final state. Revise one change at a time.
The briefs describe creative intent. Interface controls and model availability can change, so confirm the current fields in the surface you use.
The 60 canonical prompt briefs are maintained in English. Fifteen README entry languages explain the collection, workflow, category navigation, and the same complete CIN-01 example. These localized entry guides are not full translations of all 60 prompts, and their shared examples are not counted as additional prompt concepts.
[Workflow] generation, image animation, reference-led creation, editing, interpolation, or extension
[Inputs] one explicit role for every image or video
[Scene] subject, place, material, light, and stable spatial relationships
[Action] an ordered event with a visible cause and result
[Camera] framing, one primary move, and a deliberate stopping point
[Audio] ambience, action sounds, music, dialogue, or intentional silence
[Continuity] identity, geometry, object count, screen direction, and preserved regions
[Review] observable conditions that make the output acceptable or unacceptable
Gemini Omni editing prompts often benefit from a concise delta such as “change only the weather; keep everything else the same.” Generation prompts need enough scene and action detail to establish the full shot. See the prompting guide.
| IDs | Category | Briefs | Focus |
|---|---|---|---|
| CIN-01–06 | Cinematic storytelling | 6 | Single shots, reveals, transitions, visual continuity |
| ADV-01–06 | Products and advertising | 6 | Product motion, materials, demonstrations, delivery crops |
| SOC-01–06 | Creator and social video | 6 | UGC, vertical stories, tutorials, loops |
| ARC-01–06 | Architecture, travel, and place | 6 | Spatial routes, weather, scale, location continuity |
| EDU-01–06 | Education and science | 6 | Observable explanations, diagrams, processes, evidence boundaries |
| ANI-01–06 | Animation and design | 6 | Paper, clay, miniature, graphic, material transformation |
| AUD-01–06 | Dialogue, music, and sound | 6 | Speaker turns, Foley, rhythm, synchronized events |
| REF-01–06 | Reference-led creation | 6 | Identity, objects, environments, movement, mixed references |
| EDT-01–06 | Conversational video editing | 6 | Bounded changes, removals, relighting, restyling |
| EXT-01–06 | Extension and interpolation | 6 | Scene continuation, endpoint transitions, seamless loops |
CIN-01 builds one readable event around the same blue paper bird shown in the cover artwork. The prompt locks the subject, camera path, room geography, and final rooftop destination.
EDT-02 changes only a lamp body color while preserving its label-free switch, cable, reflections, camera, and timing.
EXT-03 connects two approved views of a folding exhibition wall through a physically legible movement path.
| Product interaction | Reference-led motion | Bounded weather edit |
|---|---|---|
 |
 |
 |
| ADV-01 | REF-03 | EDT-01 |
These are static concept illustrations, not Gemini Omni video outputs. See asset prompts and provenance.
- Use the use-case matrix to choose a workflow from the inputs you already have.
- Use troubleshooting to isolate motion, identity, edit-boundary, sound, text, interpolation, and extension problems.
Google documents gemini-omni-1.1-flash as a video model that accepts text, image, and video input and produces video. Current documentation covers conversational editing, image-to-video creation, first-and-last-frame interpolation, reference media, video extension, and resolution workflows. Review the official pages for current availability and request details:
- Gemini Omni Flash model page
- Generate and edit videos with Gemini Omni Flash
- Gemini Omni Flash model card
Google also notes that complete consistency through edits, complex motion, and perfectly accurate text can remain challenging. Treat every instruction in this library as a production target to inspect, not a guaranteed result.
bestimage.ai maintains this prompt collection alongside its image and video generation platform. Browse the current API documentation and model catalog for available workflows, credentials, request fields, and pricing before integrating a model.
Read CONTRIBUTING.md before submitting a prompt or media example. Contributions should state the exact workflow, input roles, complete prompt, settings, rights, and observable result.
This prompt library is curated and maintained by the bestimage.ai team, connecting practical creative workflows with image and video model APIs.
Build tutorials, share prompts, or publish API integrations? Join the bestimage.ai Affiliate Program and earn commissions by introducing your audience to bestimage.ai.
- 20% on a referred user's first valid paid order.
- 10% on subsequent valid paid orders made within 60 days after that user registers.
Order eligibility and payouts follow the current affiliate agreement.
MIT.