Kling O1 is a video editor you drive with sentences. Upload a clip, write "change the jacket to red and make it late afternoon," and O1 applies it — while the camera move, the body movement, and the timing of the original stay exactly as they were.
No timeline. No masks. No keyframes. That is the whole O1 interface.
This guide covers what Kling O1 edits well, the prompt syntax that makes it precise, what it costs against the alternatives, and the jobs where a different model is the right call. Last updated July 2026.
TL;DR — Kling O1 in one box
- What it is: a multimodal model built specifically for editing existing footage with natural-language instructions.
- What it preserves: camera paths, body movement, and timing from the source clip. O1 analyzes spatial structure and timing before applying your edit.
- Input: MP4, MOV, WebM, M4V, or GIF — 3 to 10 seconds, max 200MB, 720px to 2160px. Plus up to 7 reference images.
- Cost: 20 credits/second in Standard without a video input, 30/second with one; 25 and 40 in Pro. A 10-second edit is 300 credits.
- The syntax worth knowing: write
<<<image_1>>>and<<<video_1>>>directly in the prompt to point at specific uploads.
See the exact credit cost before you render → open the generator
What Kling O1 actually edits
O1 is built around one idea: your footage is the ground truth, and the edit happens on top of it. O1 reads the 3D structure and the timing of the clip first, then applies changes without rebuilding the motion.
Four jobs O1 is explicitly designed for:
- Wardrobe and appearance. Change outfits, adjust lighting, alter weather in existing footage.
- Style transfer. Apply a visual style while the original camera work and motion survive intact.
- Character replacement. Swap a character using reference images while their movements and interactions are preserved.
- Environment changes. Change season, time of day, or location while the action continues unchanged.
That third one is worth pausing on, because it answers a question people ask constantly and usually get pointed at the wrong tool for. If you want to replace the person in a clip and keep that clip — its background, its framing, its lighting — that is Kling O1. If you want your character performing the movement from a reference video in a newly generated scene, that is Motion Control, which rebuilds the environment rather than preserving it.
The Kling O1 reference syntax
This is the part that separates a vague edit from a precise one.
Inside your prompt, you can point at uploaded media by name:
<<<video_1>>>refers to your uploaded video<<<image_1>>>refers to your first reference image
So instead of hoping "the woman" resolves to the right subject, you write an instruction that names its inputs directly — replace the person in <<<video_1>>> with the character in <<<image_1>>>, keep the walk and the camera move. With up to 7 reference images available, that addressing scheme is what keeps a multi-reference edit unambiguous.
Kling O1 Feature mode vs Base mode
Kling O1 has two reference modes, and picking the wrong one wastes the generation:
- Base mode — direct editing with full motion preservation. The source clip is the foundation; your prompt changes what things look like, not how they move. This is the mode for actual editing.
- Feature mode — style guidance. The reference informs the look while new motion is generated.
If your goal is "this clip, but different," you want O1 in Base mode. Feature is for borrowing an aesthetic.
What Kling O1 costs
Kling O1 prices by whether you feed it a video. Text or image-driven generation runs 20 credits/second in Standard; adding a video input raises it to 30. Pro is 25 and 40.
Kling O1 credit cost by length and mode, in credits and dollars at the annual-plan rate (~$0.008/credit).
Editing means the video-input column, so the number that matters most: a 10-second edit costs 300 credits in Standard — about $2.40 on an annual plan, or 400 credits in Pro.
Set against the alternative, that is the argument for Kling O1. The same 10-second job on Kling 3.0 Omni costs 450 credits silent, because Omni bills 45 credits/second. O1 is a third cheaper for the work it was purpose-built to do — and the model documentation draws the same line: O1 is built for editing with deeper motion analysis and preservation, while Omni is the broader generate-and-edit model.
One pricing detail worth noticing: you only pay the video-input premium when you actually use a video. Text or image-driven generation on Kling O1 stays at 20 credits/second, cheaper than Omni's silent rate.
Full plans are on the pricing page, and the credit system is explained in the Kling AI pricing guide.
Kling O1 limits to plan around
O1 input clips are 3–10 seconds. Longer footage has to be trimmed before it goes in. This is an edit-a-shot tool, not an edit-a-video tool.
200MB and 720–2160px. Generous on format — MP4, MOV, WebM, M4V, and GIF all work — but the size ceiling means high-bitrate source footage may need a pass through compression first.
O1 reference images cap at 7. Enough for a character from several angles, not enough for a full style library.
O1 edits, it does not extend. O1 changes what is in your clip. It will not add seconds to the end or generate a follow-on shot.
Which model for which job
| Your goal | Model |
|---|---|
| Change something in footage you already have | Kling O1, in Base mode |
| Generate new video from references, with audio | Kling 3.0 Omni |
| Put your character through a motion from another clip | Motion Control |
| A portrait speaking your audio | Kling Avatar V2 |
The honest test for Kling O1: if you can point at an existing clip and finish the sentence "same shot, but ___," O1 is the model. If you cannot, something else is cheaper.
FAQ
What is Kling O1?
Kling O1 is a multimodal AI model built for video editing with natural language. You upload a clip of 3–10 seconds, describe the change in plain text, and Kling O1 applies it while preserving the original camera paths, body movement, and timing. No timeline, masks, or keyframing are involved.
How much does Kling O1 cost?
20 credits per second in Standard mode without a video input and 30 per second with one; 25 and 40 per second in Pro. A 10-second edit therefore costs 300 credits in Standard or 400 in Pro, and a 5-second edit costs 150 or 200. You only pay the video-input premium when you actually supply a video.
How is Kling O1 different from Kling 3.0 Omni?
O1 is purpose-built for editing, with deeper motion analysis and preservation; Omni is a broader model covering both generation and editing. O1 is also cheaper for editing — 30 credits/second with video input against Omni's 45 — so use O1 when the job is precise changes to footage you already have.
What is the <<<video_1>>> syntax in Kling O1?
It is reference syntax for prompts. Writing <<<video_1>>> or <<<image_1>>> inside your instruction points at a specific uploaded file, so a multi-reference edit stays unambiguous — for example, replacing the person in <<<video_1>>> with the character shown in <<<image_1>>>.
Can Kling O1 replace a character in a video?
Yes. Character replacement is one of its designed use cases: supply reference images of the replacement character, and Kling O1 swaps them in while preserving the original movements and interactions. Unlike motion transfer, the source clip's background and framing are kept.
What video formats and lengths does Kling O1 accept?
MP4, MOV, WebM, M4V, and GIF, from 3 to 10 seconds, up to 200MB, at resolutions between 720px and 2160px. Longer source footage must be trimmed to a single shot before editing.
What is the difference between Base and Feature mode?
Base mode is direct editing with full motion preservation — the source clip is the foundation and the prompt changes appearance only. Feature mode uses the reference as style guidance while generating new motion. For editing an existing clip, Base is the correct mode.
Try Kling O1 with free credits
New accounts start with 100 free credits, and the generator shows the exact O1 estimate before every render, so you can price a Kling O1 edit before committing. Failed generations are refunded automatically.
For the models next to this one, see Free AI Talking Avatar Generator, Kling 3.0 Turbo Camcorder Prompts and Kling 4.0 Release Date.
Edit a clip with Kling O1 → · See the model specs → · Compare plans and credits →
