Gemini Omni — natively multimodal any-to-any model. Generates high-fidelity video with synchronized audio directly from text prompts, with unified reasoning across modalities for more coherent scenes and fewer pipeline artifacts.
About this model
Gemini Omni is Google's first natively multimodal any-to-any foundation model, unveiled at I/O 2026. Unlike pipelines that chain specialized systems, Gemini Omni reasons across text, image, audio, and video in a single forward pass — producing high-fidelity video with synchronized audio directly from a single prompt.
The first model in the family, Gemini Omni Flash, focuses on cinematic video generation with coherent motion, realistic physics, and natively generated dialogue and ambient sound — eliminating the seams typical of multi-stage generative stacks.
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapi | $0.90–$1.80 (720p/1080p) · $2.10–$3.00 (4K) | Price scales with duration (4–10 s) and resolution. Synchronized audio included at no extra charge. |
| Fal.ai | Not available | Gemini Omni is not currently available on Fal.ai. |
| Replicate | Not available | Gemini Omni is not currently available on Replicate. |
Price scales with duration (4–10 s) and resolution. Synchronized audio included at no extra charge.
Gemini Omni is not currently available on Fal.ai.
Gemini Omni is not currently available 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. Gemini Omni supports rich multimodal prompts including scene composition, camera direction, dialogue, and ambient audio cues. | A clock begins ticking louder and rapidly grows larger, breaking through the table and floor. Its gears spin violently as the hands rotate uncontrollably. Walls crack apart as the giant clock expands until it fills the entire apartment. |
| Duration (seconds) | Enum (4 options) | Duration of the generated video in seconds. | 8 |
| Resolution | Enum (3 options) | Output video resolution. 720p and 1080p are the same price; 4K costs more. | 1080p |
| Aspect Ratio | Enum (2 options) | Output video aspect ratio. | 16:9 |
| Audio IDs | array | Up to 3 voice profile IDs returned by the Gemini Omni Audio endpoint. | - |
| Seed | int | Random seed (0–2147483647). Fix for reproducibility; results may still vary due to model stochasticity. | 0 |
| Character IDs | array | Up to 3 character IDs from Gemini Omni Character to feature in the video. | - |
Text description of the desired video content. Gemini Omni supports rich multimodal prompts including scene composition, camera direction, dialogue, and ambient audio cues.
A clock begins ticking louder and rapidly grows larger, breaking through the table and floor. Its gears spin violently as the hands rotate uncontrollably. Walls crack apart as the giant clock expands until it fills the entire apartment.Duration of the generated video in seconds.
8Output video resolution. 720p and 1080p are the same price; 4K costs more.
1080pOutput video aspect ratio.
16:9Up to 3 voice profile IDs returned by the Gemini Omni Audio endpoint.
-Random seed (0–2147483647). Fix for reproducibility; results may still vary due to model stochasticity.
0Up to 3 character IDs from Gemini Omni Character to feature in the video.
-Developer documentation
Write a rich multimodal prompt Gemini Omni rewards prompts that describe visuals and audio together. Include subject, action, setting, lighting, camera direction, dialogue, and ambient sound cues. Example: 'A street musician plays a violin on a rainy Paris evening, raindrops tap the cobblestones, a slow melancholic melody, distant café chatter.'
Choose duration Pick 4, 6, 8, or 10 seconds. Longer durations suit narrative scenes; shorter clips work for loops, transitions, and social posts.
Choose resolution
720p / 1080p — same price, standard quality4K — higher resolution, higher costPick an aspect ratio
16:9 — widescreen, cinematic9:16 — vertical, mobile-firstSubmit and poll
POST to /api/v1/gemini-omni-text-to-video and poll GET /api/v1/predictions/{request_id}/result until status is completed.
Prompt tips
Frequently asked
Gemini Omni is a single foundation model that reasons across text, image, audio, and video in one forward pass — instead of relaying outputs through a chain of specialized models. The result is cleaner edits, native audio generation, and fewer cross-modality artifacts.
Yes — synchronized dialogue, ambient sound, and music are generated natively in the same pass as the visuals. You can disable audio generation if you'd rather add your own soundtrack.
Gemini Omni supports 4, 6, 8, or 10 seconds per generation.
You can choose 720p, 1080p, or 4K. 720p and 1080p are priced the same; 4K costs more.
Google announced Gemini Omni at I/O 2026 on May 19, 2026, with API access rolling out in the following weeks. muapi will enable this endpoint as soon as the API is live — follow the updates page for the launch announcement.