# Edit Video Pro

> Replace a chosen span of an existing video, from 4 seconds up to 2 minutes, with new Seedance footage stitched back so the rest of the clip stays the same.

Source: https://nodaro.ai/docs/nodes/video/edit-video-pro

The **Edit Video Pro** node regenerates a chosen time window of an existing video, the replace span, and keeps the footage before and after it untouched. You set the span, describe what should happen inside it, and a Seedance 2 model generates new footage that continues from the kept footage and blends back into it. Use it to reshoot a scene, swap an action beat or restyle one moment of a longer clip.

- Found in: Video › Continue & Restyle
- Output: video
- Credits: 120–8550 per run, by model
- Models: 4
- API type: `edit-video-pro`

Edit Video Pro runs on Nodaro Cloud. On a self-hosted install, the node shows a **NODARO** mark and runs through your [connection to Nodaro Cloud](https://nodaro.ai/docs/self-hosting/cloud-connect), billed to the connected account. Without a connection, the node shows a **Connect nodaro.ai** button and cannot run.

## When to use it
- A few seconds in the middle of a clip are wrong, and the rest is good.
- An action beat should change, for example a character picks up a cup instead of a phone.
- One moment of a longer clip needs a different look.
- A long section needs new footage. Spans longer than one generation are split and stitched for you.

## Edit Video Pro, Retake Video or Extend Video

| | Edit Video Pro | [Retake Video](https://nodaro.ai/docs/nodes/video/retake-video) | [Extend Video](https://nodaro.ai/docs/nodes/video/extend-video) |
| --- | --- | --- | --- |
| **What changes** | A span inside the clip | A window inside the clip | New footage after or before the clip |
| **Shortest change** | 4 seconds | 2 seconds | Depends on the model |
| **Picture and sound** | Always both | Audio only, video only, or both | Depends on the model |
| **Models** | Seedance 2.0, 2.0 Fast, 2.0 Mini and 2.5 | LTX 2.3 Pro | Seedance 2 Extend, LTX 2.3 Pro, VEO Extend, Runway Extend |
| **Image references** | Yes | No | From the API, on Seedance 2 Extend |

## Quick start
### Add the node

Press Tab on the canvas and choose **Video › Continue & Restyle › Edit Video Pro**.

### Connect the video

Wire the source clip into the **Video** input. Open the clip's preview once, so that the span slider knows how long the clip is.

### Choose the span

Open the settings panel. Under **Replace Span**, drag the slider or type the **From (s)** and **To (s)** values. The span must be at least 4 seconds long and must stay inside the video.

### Describe the replacement

Write what happens inside the span in the prompt. Describe only the replaced window: the footage around it does not change. Wire images into **Image Refs** to show a person, a product or a place.

### Run it

Click **Run** on the node. The result is the full clip, with only the span replaced.

Workflow: Edit Video Pro replaces seconds 6 to 12 of a scene with new footage that shows a referenced umbrella, and keeps the rest of the clip.

- Upload Video → Edit Video Pro (video)
- Upload Image → Edit Video Pro (image refs)
- Text → Edit Video Pro (prompt)

## Inputs
| Input | Accepts | What it does |
| --- | --- | --- |
| **Video** | Video nodes, such as Upload Video, Generate Video and Generate Video Pro | The source clip. Required. |
| **Prompt** | Text nodes and pickers | What should replace the span. A wired prompt replaces the prompt typed in the settings panel. |
| **Image Refs** | Image nodes, such as Upload Image and Generate Image | Reference images for the new footage. The order of the connections is kept. |

The output, **Video**, is the full clip with the span replaced. It has the same structure as the source: only the requested window is different.

## Settings
| Setting | What it does |
| --- | --- |
| **Provider** | The model that generates the new footage: **Seedance 2.0** (the default), **Seedance 2.0 Fast**, **Seedance 2.0 Mini** or **Seedance 2.5**. |
| **Prompt** | What happens inside the span. |
| **Replace Span** | The window to replace, set with a slider or the **From (s)** and **To (s)** fields. The default is 0 to 8 seconds. The span must be at least 4 seconds and at most 120 seconds by default. |
| **Generate Audio** | Generates sound for the new footage. On by default. |
| **Pre & post text** | Text that is always added before and after the prompt. It is hidden from people who use your workflow as an app. See [Prompt pre and post text](https://nodaro.ai/docs/concepts/prompt-pre-post-text). |

There is no resolution or aspect ratio setting. Both come from the source video.

## Models
Edit Video Pro runs the Seedance models below. The table lists them by their catalog names, such as Seedance 2 for **Seedance 2.0**. Click a model for its settings and credit prices.

| Model | Maker | Modes | Credits | Details |
| --- | --- | --- | --- | --- |
| [Seedance 2](https://nodaro.ai/docs/models/video/seedance-2) | Bytedance | Image to video, Text to video | from 230 | Seedance 2 — premium tier with native audio. Per-second pricing by resolution. |
| [Seedance 2 Fast](https://nodaro.ai/docs/models/video/seedance-2-fast) | Bytedance | Image to video, Text to video | from 180 | Cheaper / quicker Seedance 2 tier. |
| [Seedance 2 Mini](https://nodaro.ai/docs/models/video/seedance-2-mini) | Bytedance | Image to video, Text to video | from 120 | Budget Seedance 2 tier — 480p/720p only, per-second pricing by resolution. |
| [Seedance 2.5](https://nodaro.ai/docs/models/video/seedance-2-5) | Bytedance | Image to video, Text to video | from 340 | Seedance 2.5 — up to 30s in one shot, native audio, wide multimodal references. 480p/720p/1080p. |

### Which resolution the new footage uses

The new footage uses the highest resolution tier of the chosen model that is not larger than the shorter side of the source video.

| Model | Resolution tiers |
| --- | --- |
| **Seedance 2.0** | `480p`, `720p`, `1080p` and `4K` |
| **Seedance 2.0 Fast** | `480p` and `720p` |
| **Seedance 2.0 Mini** | `480p` and `720p` |
| **Seedance 2.5** | `480p`, `720p` and `1080p` |

For example, a 4K source is bridged at 4K with Seedance 2.0, and at `720p` with Seedance 2.0 Fast, because Fast has no higher tier.

## How the replacement works

1. **The cut points become references.** When at least 2 seconds of original footage come before the span, that footage is a head reference, so the new footage continues naturally from it. When at least 2 seconds come after the span, that footage is a tail reference, so the new footage blends into what comes next. A span that starts at 0:00 has no head reference. A span that ends at the end of the video has no tail reference.
2. **The model generates the bridge.** The chosen Seedance model generates the replacement, anchored by the references that exist. When the span plus a short stitch buffer is longer than one generation, which is 15 seconds, the bridge is split into segments. Each segment continues from the last frames of the one before, as in [Generate Video Pro](https://nodaro.ai/docs/nodes/video/generate-video-pro).
3. **The pieces are stitched back.** The kept footage before and after the span is cut and joined with the new bridge into one continuous video.

## Span rules

- **The shortest span is 4 seconds.**
- **The longest span is 120 seconds by default.** On a self-hosted install, an administrator can change the cap with the `EDIT_VIDEO_PRO_MAX_SPAN` environment variable. `GET /v1/nodes` reports the active cap.
- **The span must fall inside the video.** Keep some original footage on at least one side. A span that covers the whole clip has nothing to anchor to, so use [Generate Video Pro](https://nodaro.ai/docs/nodes/video/generate-video-pro) or [Generate Video](https://nodaro.ai/docs/nodes/video/generate-video) instead.
- **A span past the end of the video is refused.** The node checks the real length of the source before it reserves any credits.

## Credits
Every run costs a flat fee plus a per-second rate for the new footage and its reference clips:

`credits = 100 + ceil(rate × (bridge seconds + reference seconds))`

- **100** is a flat fee for measuring, bridging and stitching, charged once per run.
- **rate** is the model's per-second rate for generation with a reference video, at the resolution of the bridge. For Seedance 2.0 at `720p`, the rate is 62.5 credits per second.
- **bridge seconds** is the replaced span, plus 0.3 seconds for each kept edge to cover the stitch, rounded to whole seconds.
- **reference seconds** is 2 seconds for each reference clip used, before and after the span, plus 2 seconds for each join between segments of a long bridge.

| Example (Seedance 2.0, `720p`) | Span | Source | Segments | Credits |
| --- | --- | --- | --- | --- |
| Shortest span, in the middle | 4 s | 20 s | 1 | 663 |
| Span in the middle | 10 s | 20 s | 1 | 1,038 |
| Span that starts at 0:00 | 10 s | 20 s | 1 | 850 |
| Span that ends at the end of the video | 10 s | 20 s | 1 | 850 |
| Longer span, in the middle | 20 s | 40 s | 2 | 1,850 |

A span in the middle keeps two reference clips, so it costs more than the same span at either end of the video. Nodaro measures the source video before it reserves credits, so the reservation normally equals the final charge. If the source cannot be measured, the reservation assumes the model's highest resolution, and the difference is refunded when the run completes.

## What to expect from the result

- **The output plays at 24 fps.** Every Edit Video Pro result has this frame rate.
- **The kept footage is re-encoded.** It is cut with a high-quality encode and joined to the new footage, so the result is one consistently encoded video, not a byte-for-byte copy of the original.
- **Free-plan results carry the standard watermark**, as on every node that outputs video.
- **A failed or cancelled run is refunded in full**, including the flat fee. Edit Video Pro delivers the fully stitched video or nothing.
- **Runs survive restarts.** Progress is saved after every bridge segment. If the service restarts during a run, the run resumes from the last saved segment instead of failing. A run that stalls again after resuming is failed and refunded.

## Tips
- **Leave room on both sides.** Keep at least 2 seconds of original footage before and after the span, so that the new footage has an anchor on each side.
- **Describe only the window.** The kept footage does not change, so the prompt needs to cover only what happens inside the span.
- **Use Retake Video for small fixes.** For a change shorter than 4 seconds, or one that touches only the sound or only the picture, use [Retake Video](https://nodaro.ai/docs/nodes/video/retake-video).
- **Regenerate the whole clip elsewhere.** To replace everything, use [Generate Video Pro](https://nodaro.ai/docs/nodes/video/generate-video-pro) for long clips or [Generate Video](https://nodaro.ai/docs/nodes/video/generate-video) for single shots.
- **Add footage at the end instead.** To make a clip longer, use [Extend Video](https://nodaro.ai/docs/nodes/video/extend-video).

## Frequently asked questions

### How long can the replaced span be in Edit Video Pro?

At least 4 seconds and at most 120 seconds by default. Spans longer than about 15 seconds are split into several Seedance segments and stitched together automatically.

### Does Edit Video Pro work on a self-hosted install?

Yes, through a connection to Nodaro Cloud, and the runs are billed to the connected account. Without a connection, the node shows a Connect nodaro.ai button and cannot run.

### What happens to my credits if an Edit Video Pro run fails?

Every failure or cancellation refunds the full reservation, including the flat fee. The node never delivers a half-replaced span.

### Why does a span in the middle of a clip cost more than a span at the start?

A span in the middle uses two reference clips, one before and one after the span, and each is billed as 2 seconds. A span that starts at 0:00 or ends at the end of the video uses only one.

### Can I choose the resolution of the new footage?

No. Nodaro takes the resolution from the source video. The new footage uses the highest resolution the chosen model supports that is not larger than the source.