Kling 3.0 Omni Image-to-Video and Kling 3.0 Omni Text-to-Video are the same engine at the same price — identical 3-to-15-second range, identical 720p/1080p/4K tiers, identical per-second credit table — and the only real differences are that image-to-video starts from a photo you upload and inherits that photo’s aspect ratio, while text-to-video starts from words alone and lets you choose 16:9, 9:16 or 1:1. Pick by what you already have in hand, not by quality, because the quality is the same model.

Why the pricing is identical

On VdoBloom both modes read from one credit table. There is no premium for uploading an image and no discount for going text-only. That is unusual — many families charge differently for the two paths — and it means you can switch modes mid-project without re-planning a budget.

Length	720p credits	1080p credits	4K credits	1080p dollar cost
3s	24	33	81	$1.65
4s	32	44	108	$2.20
5s	40	54	134	$2.70
6s	48	65	161	$3.25
8s	64	87	215	$4.35
10s	80	108	268	$5.40
12s	96	130	322	$6.50
15s	120	162	402	$8.10

Those tiers work out at 8 credits per second on 720p, 10.8 on 1080p and 26.8 on 4K. One credit is $0.05 at Lite plan rates ($15 for 300 credits), so a 10-second 720p clip is $4.00 in either mode. Annual billing brings Lite to $10 a month, which lowers the effective cost of every row above.

The differences that actually matter

Feature	Omni Image-to-Video	Omni Text-to-Video
Starting input	One uploaded photo plus a prompt	Prompt only
Aspect ratio control	None — inherits the image, sent as auto	16:9, 9:16 or 1:1 picker
Durations	3-15s, every second	3-15s, every second
Resolutions	720p, 1080p, 4K	720p, 1080p, 4K
Native audio	Always on	Always on
Dashboard tab	Image to Video	Text to Video
Subject consistency	Locked to your photo	Model invents the subject
Credits per second	8 / 10.8 / 26.8	8 / 10.8 / 26.8

The aspect-ratio behaviour is the one that trips people up. Image-to-video deliberately hides the ratio picker because the single-shot path requires an automatic ratio — the output follows the frame of the photo you upload. If you need a 9:16 vertical from image-to-video, crop the source photo to 9:16 first. Text-to-video has no source frame to inherit, so it exposes the picker and defaults to 16:9.

Pick image-to-video when the subject already exists

Use image-to-video whenever the look is non-negotiable: a specific product, a specific face, a brand asset, a photo you already shot and approved. The model animates what you gave it rather than approximating it, which removes the single most expensive failure mode in text-to-video — a beautiful clip of the wrong thing.

A worked example. You have an approved packshot of a bottle and want a 6-second 1080p hero loop with ambient room sound. Open the Image to Video tab, pick Kling 3.0 Omni Image-to-Video, upload the packshot cropped to 16:9, set 6 seconds and 1080p, and prompt for a slow push-in with soft light drifting across the label. That is 65 credits, or $3.25, and the bottle looks exactly like your bottle.

If the photo shows a real person, you must have that person’s consent before animating them. That applies to every photo-driven model on VdoBloom, not just this one.

Pick text-to-video when the shot does not exist yet

Text-to-video is the faster path for establishing shots, b-roll, abstract transitions and anything you would otherwise have to source a stock photo for just to feed the image path. It is also the only one of the two where you can hold a vertical 9:16 format without preparing an asset, which matters when you are producing for short-form feeds at volume.

The trade is control. You are describing a subject rather than supplying one, so two generations of the same prompt will not share a face, a garment or a logo. For a campaign that needs the same character across four clips, text-to-video will fight you; generate one still you like, then drive all four clips from it through image-to-video.

Prompting each mode

For image-to-video, prompt the motion and the sound, not the scene. The scene is in the photo. Write what moves, how the camera behaves and what you should hear: a slow dolly-in, hair lifting in a breeze, distant traffic. Describing what is already visible wastes prompt weight and sometimes fights the source frame.

For text-to-video, prompt the scene first and the motion second, and be concrete about time of day, lens feel and location. Because Omni generates audio natively, naming the soundscape in the prompt genuinely changes the output.

Both modes sit in the Kling family, which carries a strict content-filter label on VdoBloom. Swimwear, dance and fitness briefs that Kling declines should be routed to a flexible-tier model such as Wan 2.7 Image-to-Video. Explicit and illegal content is blocked across the whole platform.

When to use neither

If you are running ten drafts to find a direction, 3-second 720p Omni clips at 24 credits each are the cheap way to do it, but a fixed-length model like Kling 2.6 Image-to-Video at 22 credits for a full 5 seconds is cheaper still for silent drafting. Save Omni for the take you intend to ship.

Full specs for both modes are on the Kling 3.0 Omni Image-to-Video page and the Kling 3.0 Omni Text-to-Video page, and every plan and credit pack is listed on pricing. New accounts get 10 free credits with no card, enough for a 3-second 720p test on either mode once you top up slightly, and one-time packs start at $2.49 for 75 credits that never expire.

Frequently asked questions

Does image-to-video cost more than text-to-video on Kling 3.0 Omni?

No. Both modes bill from the same table: 8 credits per second at 720p, 10.8 at 1080p and 26.8 at 4K.

Why can I not pick an aspect ratio on the image-to-video mode?

Single-shot image-to-video takes its ratio from the uploaded photo, so the picker is hidden to stop a leftover ratio from another model being sent. Crop the photo to the ratio you want.

Can I get a 9:16 vertical out of Kling 3.0 Omni?

Yes, both ways. Choose 9:16 in text-to-video, or upload a 9:16-cropped photo to image-to-video.

Is the audio the same in both modes?

Yes. Native audio is generated with the video in both modes and cannot be switched off to save credits.

What is the cheapest way to test both?

Run one 3-second 720p clip in each mode: 24 credits each, 48 credits total, about $2.40 at Lite plan rates.

Kling 3.0 Omni Image-to-Video vs Text-to-Video: Same Price, Different Job