Picsart now carries two Gemini Omni video models, and 1.1 Flash is a bigger change than its version number suggests. The first Gemini Omni built a shot from a description. Version 1.1 takes references, keyframes, and existing footage, and it outputs up to 4K.

So the choice between them is not really about quality. Pick the original when a written prompt is the whole brief. Pick 1.1 when you need to steer the shot toward something specific, or when the file has to hold up past 720p.

Gemini Omni 1.1 Flash vs Gemini Omni at a glance

Capability Gemini Omni Gemini Omni 1.1 Flash
Output resolution Up to 720p Up to 4K
Reference images per generation One Up to five
Reference videos per generation One Up to three
Start and end frame Not available Both, set independently
Clip extension Not available Available
Clip lengths 3, 5, 6, 8, or 10 seconds Any whole second from 3 to 10
Aspect ratios 16:9 and 9:16 16:9 and 9:16
Video plus synchronized audio One pass One pass
Audio file as input Not accepted Not accepted
Chat-edit after generating Yes Yes

Read down that table and the shape of the update is clear. Version 1.1 is not simply sharper. It has more inputs, and every one of them is a way to pin down what comes back.

What the extra inputs let you control

Five reference images instead of one. A single shot can carry a product, a person, a setting, and a look at once. There is no need to composite the ingredients into one image before prompting.

Three reference videos instead of one. Motion becomes reference material rather than something described in words. Feed the model the camera move you want alongside the performance you want, and ask for a shot that honors both.

A start frame and an end frame. The original has no keyframe slots at all. Version 1.1 takes both ends of a move as images and generates the frames between them, which lands a specific camera move far more reliably than describing it.

Clip extension. Existing footage goes in and the model carries the shot on instead of starting a new one. That is the piece that makes Omni useful on projects that began somewhere else.

Every whole second from three to ten. The original offers three, five, six, eight, and ten. Version 1.1 fills in four, seven, and nine, which matters when a cut needs an exact length and the alternative is generating longer and trimming.

What 4K actually buys you

Reframing stops being destructive. A 4K master holds enough pixels to crop a 16:9 shot to a 9:16 vertical and still deliver a sharp full-HD cut. At 720p that crop leaves nowhere near enough frame to do it.

The footage survives a real edit. Color grading, stabilization, and speed ramps all cost detail. Starting at 4K leaves headroom for those passes, so the version reaching the timeline is not already the version with nothing left to give.

It clears the bar for work that leaves a feed. Anything projected, shown on a large display, cut into a broadcast package, or pulled as a still for print needs more than 720p. The original cannot serve any of it without a separate upscaling pass.

When the original Gemini Omni is still enough

The first version does not become obsolete, and picking it deliberately is sometimes the better call.

A prompt-only brief. No references, no keyframes, nothing to continue. The extra slots on 1.1 sit unused, and the original covers the job.

A social-only deliverable. A vertical cut destined for one feed gets compressed on upload regardless. Generating above the delivery spec changes nothing a viewer will see.

Fast iteration on an idea. Testing ten variations of a concept costs less at a lower resolution. Settle the composition first, then generate the keeper on 1.1.

Which Gemini Omni to pick

  • A 4K master, or any work that leaves a social feed. Gemini Omni 1.1 Flash.
  • One generation serving both a landscape and a vertical placement. Version 1.1, then crop from the 4K frame.
  • A camera move you can draw but would struggle to describe. Version 1.1, with a start frame and an end frame.
  • A shot that has to match a product, a person, or an existing look. Version 1.1, with up to five reference images.
  • Footage you already shot and want continued. Version 1.1, which is the version with extension.
  • A cut that has to be exactly four or seven seconds. Version 1.1.
  • A vertical clip generated from a written prompt alone. Either one. Gemini Omni is the leaner pick.

Try Gemini Omni 1.1 Flash on one shot

The upgrade is easiest to judge on a shot you intend to crop. Generate this at 16:9, then pull a 9:16 vertical out of the same frame and look at how the vertical holds up.

Reframe and crop test

Slow handheld push-in on a ceramic coffee cup on a weathered wooden table, steam rising, a blurred cafe window behind it. Start wide enough to show the full table and end tight on the rim of the cup. Warm late-afternoon light raking in from the left, soft shadow falling to the right, shallow depth of field, fine grain on the wood grain and the glaze. Quiet room tone, a distant espresso machine, one soft ceramic clink as the camera settles

Why run Gemini Omni on Picsart

Gemini Omni is Google’s model, and Google puts it in plenty of places. The reason to run it here is that both versions sit in the same Picsart AI Playground picker, so trying one against the other is a dropdown and the prompt stays where it is.

That picker holds 176 AI models from 32 providers on one credit balance, with each model’s cost shown before you generate. Nothing idles on a subscription for a model you reach for twice a month. The Google Omni model page covers the family, and the video models lineup covers what else suits a given shot.

Get answers to common questions

Inputs and resolution. Version 1.1 adds up to five reference images, three reference videos, an independent start and end frame, and clip extension, and it outputs up to 4K where the original stops at 720p.

Generate your first 4K Gemini Omni clip

Pick a shot you plan to crop two ways, generate it on 1.1, and see how much the vertical gains. Start in Picsart AI Playground.