Generate video from text with Gemini Omni 1.1 Flash. Add reference images, a reference video, voice, and character IDs in one call with native audio. Try free.
Gemini Omni 1.1 Flash — text-to-video generation with native synchronized audio, supporting optional reference images, a reference video clip, voice profile, and character reference inputs in one call.
About this model
Gemini Omni 1.1 Flash is Gemini Omni's text-to-video endpoint: write a prompt and optionally layer on up to 7 reference images (auxiliary conditioning, not a starting keyframe), a short reference video clip, up to 3 voice profiles from Gemini Omni Audio, and up to 3 character references from Gemini Omni Character. The model reasons jointly across whichever inputs you provide and returns a video with natively synchronized dialogue, ambient sound, and music in the same generation pass — no separate audio pipeline required. To animate a specific starting image (and optionally an ending image) instead of generating from a prompt alone, use Gemini Omni 1.1 Flash Image to Video. For editing an existing video, see Gemini Omni 1.1 Flash Video Edit; for tag-addressed multi-reference composition, see Gemini Omni 1.1 Flash Reference to Video.
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapi | $0.10–$0.30 per second of output (by resolution) | Pay-per-generation, no subscription. Covers text-to-video plus optional image, video-reference, voice, and character inputs. |
| Fal.ai | Not available | Gemini Omni 1.1 Flash is not listed on Fal.ai. |
| Replicate | Not available | Gemini Omni 1.1 Flash is not listed on Replicate. |
Pay-per-generation, no subscription. Covers text-to-video plus optional image, video-reference, voice, and character inputs.
Gemini Omni 1.1 Flash is not listed on Fal.ai.
Gemini Omni 1.1 Flash is not listed on Replicate.
** Competitor pricing is estimated based on similar model architectures and usage tiers.
Configuration schema
| Parameter | Type | Description | Default |
|---|---|---|---|
| Prompt | string | Text description of the desired video content — visuals, camera direction, dialogue, and ambient audio cues. | A street musician plays a violin on a rainy Paris evening, raindrops tap the cobblestones, a slow melancholic melody, distant café chatter. |
| Reference Images | array | Up to 7 reference images used as auxiliary conditioning (not a starting keyframe). Each counts as 1 quota unit against the shared 7-unit total with video and character_ids. | - |
| Reference Video | string | A reference video clip, max 100MB / 30s. Counts as 2 quota units. | - |
| Video Trim Start (s) | number | Start time, in seconds, of the reference video window. | 0 |
| Video Trim End (s) | number | End time, in seconds, of the reference video window. Window must not exceed 10 seconds. | 10 |
| Audio IDs | array | Up to 3 voice profile IDs from the Gemini Omni Audio endpoint. Each counts as 1 quota unit. | - |
| Character IDs | array | Up to 3 character IDs from Gemini Omni Character. Each counts as 1 quota unit. | - |
| Duration (seconds) | Enum (4 options) | Duration of the generated video in seconds. Ignored when a reference video is provided — output duration is then determined by the model. | 8 |
| Aspect Ratio | Enum (2 options) | Output video aspect ratio. | 16:9 |
| Resolution | Enum (3 options) | Output resolution. Billed per second of output: $0.10/s at 720p, $0.15/s at 1080p, $0.30/s at 4K. | 720p |
| Seed | int | Random seed (0–2147483647). Fix for reproducibility; results may still vary due to model stochasticity. | 0 |
Text description of the desired video content — visuals, camera direction, dialogue, and ambient audio cues.
A street musician plays a violin on a rainy Paris evening, raindrops tap the cobblestones, a slow melancholic melody, distant café chatter.Up to 7 reference images used as auxiliary conditioning (not a starting keyframe). Each counts as 1 quota unit against the shared 7-unit total with video and character_ids.
-A reference video clip, max 100MB / 30s. Counts as 2 quota units.
-Start time, in seconds, of the reference video window.
0End time, in seconds, of the reference video window. Window must not exceed 10 seconds.
10Up to 3 voice profile IDs from the Gemini Omni Audio endpoint. Each counts as 1 quota unit.
-Up to 3 character IDs from Gemini Omni Character. Each counts as 1 quota unit.
-Duration of the generated video in seconds. Ignored when a reference video is provided — output duration is then determined by the model.
8Output video aspect ratio.
16:9Output resolution. Billed per second of output: $0.10/s at 720p, $0.15/s at 1080p, $0.30/s at 4K.
720pRandom seed (0–2147483647). Fix for reproducibility; results may still vary due to model stochasticity.
0Developer documentation
Write a multimodal prompt Describe the visuals, camera direction, dialogue, and ambient sound together. Example: 'A street musician plays a violin on a rainy Paris evening, raindrops tap the cobblestones, a slow melancholic melody, distant café chatter.'
Layer on optional conditioning
image_urls, up to 7) as auxiliary visual conditioning.video_url + trim_start/trim_end, max 10-second window) for continuity-driven extension.audio_ids) and/or character references (character_ids) layered on top.Respect the shared quota Images count 1 unit each, the reference video counts 2 units, and each character ID counts 1 unit — the total must not exceed 7.
Pick duration, resolution, and aspect ratio Duration (4/6/8/10s) is ignored when a reference video is supplied — the model determines output length. Resolution is billed per second of output: $0.10/s at 720p, $0.15/s at 1080p, $0.30/s at 4K.
Submit and poll
curl -X POST https://api.muapi.ai/api/v1/gemini-omni-flash-1-1-text-to-video \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "A street musician plays a violin on a rainy Paris evening",
"image_urls": ["https://example.com/scene.jpg"],
"duration": 8,
"resolution": "1080p",
"aspect_ratio": "16:9"
}'
Then poll GET /api/v1/predictions/{request_id}/result until status is completed.
Prompt tips
Frequently asked
It's Gemini Omni's text-to-video endpoint: it accepts a prompt plus optional reference images, a reference video clip, voice profiles, and character IDs in a single call, and returns video with natively synchronized audio.
Yes. Reference images, a reference video, voice profiles, and character IDs can all be combined, subject to the 7-unit shared quota.
Each reference image consumes 1 unit, the reference video consumes 2 units, and each character ID consumes 1 unit. The total across images + (video × 2) + character IDs must not exceed 7.
It's ignored. When a reference video is supplied, the model determines the output duration automatically instead of using the requested `duration` value.
Billed per second of output video by resolution: $0.10/s at 720p, $0.15/s at 1080p, and $0.30/s at 4K. A 10-second 1080p clip costs $1.50.
Generate reusable voice profiles with [Gemini Omni Audio](/playground/gemini-omni-audio) and reusable character references with [Gemini Omni Character](/playground/gemini-omni-character), then pass their returned IDs here.
No. Use [Gemini Omni 1.1 Flash Image to Video](/playground/gemini-omni-flash-1-1-image-to-video) instead — its `first_frame_url` is the actual starting keyframe. This endpoint's `image_urls` are auxiliary reference images, not a keyframe.