Gemini Omni 1.1 Flash API — Keyframes, Video Reference & 4K

Live1 unified endpoint

Google announced Gemini Omni 1.1 Flash on August 27, 2026 — an update to Gemini Omni Flash adding first/last frame keyframe control, a short reference-video input, and 4K upscaling. Muapi wraps it as a single unified API endpoint that combines text, images, keyframes, video reference, voice, and character inputs in one call.

Gemini Omni 1.1 Flash is available now via API

Try it in the playground or the original single-purpose Gemini Omni Flash endpoints.

Live
T2V

Gemini Omni 1.1 Flash — Text & Image to Video

Generate video from a text prompt alone, or animate up to 7 reference images, with native synchronized audio in the same pass.

720p–4K
Try Gemini Omni 1.1 Flash Now
Live
Keyframe

Gemini Omni 1.1 Flash — Keyframe Control

Specify a first and last frame and generate the connecting motion between them — new in the 1.1 update.

720p–4K
Try Gemini Omni 1.1 Flash Now
Live
V2V

Gemini Omni 1.1 Flash — Video Edit

Edit a source video with a single text instruction — restyle, relight, or transform footage while preserving the original motion and timing.

360p–4K
Try Gemini Omni 1.1 Flash Now

What is Gemini Omni 1.1 Flash?

Gemini Omni 1.1 Flash is Google's update to Gemini Omni Flash, the natively multimodal any-to-any model first announced at I/O 2026. Where the original Flash model focused on generating a single clip with synchronized audio in one forward pass, 1.1 adds the production controls creators actually need for more precise scenes: keyframe control via first/last frame specification, a reference-video input for style and motion continuity, and 4K upscaling for final delivery.

Muapi wraps Gemini Omni 1.1 Flash as a single unified submit-and-poll REST endpoint that accepts text, reference images, keyframes, a reference video clip, voice profiles, and character IDs in one call.

What's New in 1.1

First/Last Frame KeyframesSpecify a start and end frame and generate continuous, smooth video that transitions between them.
4K UpscalingUpscale final output to 4K for delivery, on top of the existing 720p and 1080p tiers.
Video Reference InputReference up to a 10-second window of an existing video clip when crafting a new scene, for style and motion continuity.
Native Synchronized AudioCarries forward Gemini Omni Flash's native audio generation — dialogue, ambient sound, and music — in the same forward pass.
One Unified EndpointText, images, keyframes, video reference, voice, and character IDs all go through a single request — no separate endpoint per input mode.

Use Cases

🔑Keyframed TransitionsAnchor first and last frames for precise, controllable transitions between two shots.
📼Style-Consistent SequelsReference up to a 10-second window of an existing clip to keep a new scene visually and tonally aligned.
🎭Consistent Characters & VoicesLayer saved character and voice profiles onto images or a reference video for recurring-cast scenes.
🗣️Multimodal Narrative ShortsDrive scene composition, camera direction, and dialogue from one rich text prompt with native audio.
📺Final Delivery at 4KShip broadcast- and streaming-ready output once a draft has been approved.

Gemini Omni 1.1 Flash vs Gemini Omni Flash — What's New

FeatureGemini Omni 1.1 FlashGemini Omni Flash
Keyframe ControlFirst/last frame specificationNot supported
Video Reference InputUp to a 10-second windowNot supported
Resolution Options720p, 1080p, 4K upscale720p, 1080p, 4K
Endpoint ShapeOne unified endpoint for all input modesSeparate endpoint per input mode
StatusAvailable now on MuapiAvailable now on Muapi

Frequently Asked Questions

What is Gemini Omni 1.1 Flash?

Gemini Omni 1.1 Flash is Google's update to Gemini Omni Flash, announced August 27, 2026. It adds first/last frame keyframe control, a reference-video input, and 4K upscaling, all through one unified endpoint that also accepts text, images, voice profiles, and character IDs.

How is Gemini Omni 1.1 Flash different from the original Gemini Omni Flash?

The 1.1 update lets you specify first and last frames for keyframe-controlled generation, adds a reference-video input (up to a 10-second window), and adds 4K upscaling — combined into a single endpoint instead of separate ones per input mode.

Can I combine keyframes with a reference video or images?

No — `first_frame_url` (keyframe mode) is mutually exclusive with reference images, the reference video, voice profiles, and character IDs. Outside of keyframe mode, those inputs can be combined, subject to a shared 7-unit quota.

Is the Gemini Omni 1.1 Flash API available on Muapi now?

Yes — try it at /playground/gemini-omni-flash-1-1. The original single-purpose Gemini Omni Flash endpoints remain available as well.

What resolution options does Gemini Omni 1.1 Flash support?

720p and 1080p standard output, and 4K upscaling for final delivery, billed per second of output video.

Looking for Gemini Omni today?

Gemini Omni Flash is live now — text-to-video, image-to-video, video-edit, voice profiles, and character profiles are all available via API today.

Explore Gemini Omni Now