Start and End Frames
Set both endpoints of a shot so the model generates the transition between them. This is useful for controlled reveals, loops, camera orbits, and before-and-after sequences.
Loading…
Live Model Guide · Updated August 31, 2026
Turn one idea into a model-aware Gemini Omni prompt, then optionally render a 720p video with native audio inside PromptMotion. This page tracks the live 1.1 API, current pricing, supported controls, and the limits of our production integration.
Gemini Omni Flash 1.1 is a Google model. PromptMotion is an independent product and uses a third-party API provider; it is not affiliated with or endorsed by Google.
3 prompt generations free daily · Video rendering requires Pro credits
The model can do more than the controls currently exposed here. This table separates the working product from the wider Gemini Omni feature set.
| Model endpoint | google:gemini@omni-flash-1.1 |
| PromptMotion output | 720p · 16:9 or 9:16 |
| Clip length | 5 seconds or 8 seconds |
| Audio | Native audio included |
| Video credits | 30 credits (5s) · 50 credits (8s) |
| Production status | Live · end-to-end verified August 31, 2026 |
Real production check
Our August 31 test completed a 5-second 1280×720 generation end to end, returned a playable MP4 with range requests, recorded the provider cost of $0.5015415, and deducted exactly 30 video credits. Failed jobs are designed not to consume credits.
The August 27 update focuses on controllability, iteration speed, and production delivery rather than only raising visual quality.
Set both endpoints of a shot so the model generates the transition between them. This is useful for controlled reveals, loops, camera orbits, and before-and-after sequences.
Google's Gemini API can analyze more prior video context and extend a story in 10-second increments up to 40 seconds cumulative. Runware currently documents 3-to-10-second extension steps up to 30 seconds.
The 360p draft mode is designed for faster, lower-cost iteration. Selected outputs can move to 720p, 1080p, or 4K depending on the access path and workflow.
The Runware 1.1 endpoint accepts first and last frames, up to seven reference images, and up to three short reference videos for character or style guidance.
PromptMotion currently exposes the simpler 720p text-to-video path. Advanced reference, edit, extend, 1080p, and 4K controls remain on the product roadmap and are not presented as available in our interface today.
Separate prompt creation, user-facing video credits, and raw provider cost before comparing services.
Official August 27, 2026 list price. Actual platform charges can include provider fees or input processing.
| Resolution | List price | 5-second base |
|---|---|---|
| 360p draft | $0.03 / second | $0.15 |
| 720p | $0.10 / second | $0.50 |
| 1080p | $0.15 / second | $0.75 |
| 4K | $0.30 / second | $1.50 |
Prompt only
3 free / day
Build and copy a Gemini Omni-aware prompt without spending video credits or entering a credit card.
5s · 720p
30 credits
Text-to-video with native audio in 16:9 or 9:16. The connected provider cost was $0.5015415 in our latest production test.
8s · 720p
50 credits
A longer option for dialogue, camera movement, or a short narrative beat. Provider pricing and queue time can change.
Credits are PromptMotion units, not Google tokens. See the current plans and credit packs before purchasing.
PromptMotion keeps the prompt workflow consistent while letting the rendering model fit the job.
| Model | Best fit | Current in-site option |
|---|---|---|
| Gemini Omni Flash 1.1 | Fast native-audio drafts, world-aware scenes, transitions | 720p · 5s/8s · 30/50 credits |
| Seedance 2.5 Quality | Product consistency, controlled motion, quality-first output | 5s/8s · 60/100 credits |
| Seedance 2.0 Fast | Lower-cost iteration and quick visual tests | 5s/8s · 30/50 credits |
Keep one clear shot goal and write audio as part of the scene, not as an afterthought.
Name the framing and one main camera move: close-up orbit, slow dolly-in, handheld tracking, or locked wide shot.
State who or what is visible and the exact action that must complete within 5 or 8 seconds.
Give concrete spatial details instead of a generic mood: surface, background, weather, time of day, and depth.
Describe the key light, color palette, material response, and intended commercial or cinematic finish.
Write dialogue, ambience, sound effects, and music cues explicitly. Say when each sound begins or changes.
Example
Close-up orbit around a faceted amber perfume bottle on black marble. The bottle rotates once as warm gold particles spiral around the glass; crisp reflections, shallow depth of field, premium fragrance commercial, 16:9, 5 seconds. Native audio: a soft glass chime on the first turn, subtle room tone, and a restrained cinematic swell.
Gemini Omni Flash 1.1 is Google's multimodal video generation and editing model released on August 27, 2026. It accepts text and media references, produces video with audio, and adds first-and-last-frame control, scene extension, a 360p draft mode, and 1080p or 4K delivery options.
Yes. PromptMotion connects to the live google:gemini@omni-flash-1.1 endpoint through Runware. The current in-site workflow supports text-to-video at 720p, 5 or 8 seconds, and 16:9 or 9:16. It does not yet expose the model's reference-video, editing, extension, 1080p, or 4K controls.
PromptMotion includes three free model-aware prompt generations per day with no credit card. Rendering an Omni video uses Pro video credits because every generation has a real provider cost. Google may also provide limited access through its own products, subject to its current plans and regional availability.
Google's August 27 pricing table lists Gemini Omni 1.1 Flash at $0.03 per second for 360p, $0.10 for 720p, $0.15 for 1080p, and $0.30 for 4K. PromptMotion charges 30 video credits for a 5-second 720p Omni render and 50 credits for 8 seconds. In a production test on August 31, 2026, the connected provider charged $0.5015415 for one 5-second 720p result.
Yes. Google documents gemini-omni-1.1-flash through the Gemini API in Google AI Studio, while Runware exposes google:gemini@omni-flash-1.1 through its video inference API. PromptMotion uses the Runware endpoint for its current production workflow.
Describe the shot and camera, the subject and action, the setting, lighting and visual style, then the native audio or dialogue. State either 16:9 or 9:16 and keep the action achievable within the selected 5- or 8-second duration.
Describe one idea, get a Gemini Omni-aware prompt free, and decide whether the native-audio render is worth spending video credits on.
Try Gemini Omni Flash →