Ai Render Helper documentation

Start here

How Ai Render Helper works

Ai Render Helper is a Windows desktop workspace for common render-finishing tasks. Select files or folders, set the processing options, then run the selected tool. Video tools use the bundled FFmpeg/FFprobe utilities; image tools work with standard image files; the OpenPose Kit is a local 3D pose editor.

1. Download and install the Windows release from Download.
2. Open the app and use Settings to validate Pro, connect an optional ComfyUI folder, and create brand kits.
3. Choose a tool below, select its input, review the options, and start processing.
4. Keep the app open while a job runs. Video tools show progress and let you request a stop.

Desktop tools

Choose the right tool

Organizer

Use Organizer to browse one or more source roots, including configured ComfyUI folders, inspect thumbnails and previews, then send the selected item to a supported finishing tool. It is useful for sorting render attempts before trimming, branding, merging, or adding a source image to a Scenario Maker project.

Beta/developer preview

Scenario Maker

Scenario Maker is currently a beta/developer-preview feature rather than a normal public-release tab. It sends an ordered image-to-video scene queue to a configured ComfyUI installation and records run metadata and scene checkpoints.

• Scenes: add source images, reorder or duplicate scenes, use Change below a scene preview to replace its image, and add locked continuations that use the previous scene's final frame.
• WAN 2.2 settings: the current default is 1080 × 768, 121 frames, and 18 FPS. The displayed duration is frames divided by FPS. Dimensions are kept to safe multiples of 8 and WAN frame lengths use the 4n + 1 sequence.
• Recommended Settings: the preset list offers conservative 4 GB, 8 GB, 16 GB, and 24 GB profiles across horizontal, vertical, and square aspect ratios. Hardware, model precision, offload configuration, and system RAM still determine whether a render fits.
• Queue behavior: scenes run sequentially. Enable Play sound at end of queue for one completion notification; the bundled workflow's per-render sound is disabled during Scenario Maker runs.
• Recovery: every completed scene writes a checkpoint. Previous-last-frame chains are joined automatically, and Run Encode After can hand joined outputs to Encode Merge.

Encode Merge

Use this tool to turn an ordered folder of MP4 clips into one encoded output. It can crossfade each clip, loop an optional audio file, and add a branded first clip with title text and controlled camera wiggle.

Source folder required

Folder containing the MP4 clips to merge. Files are processed in name order.

Audio file optional

Audio to loop and fade into the final merged video.

Width, Height, FPS optional

Output dimensions and frame rate. Defaults are 1536 × 1536 at 24 FPS.

CrossX optional

Crossfade duration in frames. Use 0 for a sharp cut.

Truncate optional

Frames to remove from the start of every source clip before encoding.

Wiggle controls optional

Choose Low, Medium, High, Scatter, or Custom. %Wiggle controls motion strength; S.Wiggle controls its duration; Flow keeps movement continuous and Scatter changes position between segments.

Brand, Intro, Title optional

Select None or a brand. Overlay on first brands the first clip; Duplicate first creates a branded intro before it. Title text appears on the branded intro.

Process starts the job. While a job is running, the same button requests a stop and temporary files are cleaned up.

Duplicate Speed Up

Create a numbered MP4 series, add cropped first/last variants, then generate progressively faster duplicate rounds. The tool writes a rename map so Revert can restore the original naming and remove generated files when possible.

Folder required

Folder containing MP4 videos. The first video is shown in the preview.

Duplicates optional

Number of duplicate rounds to create. Free mode is limited to one round.

Speed up optional

Percentage added progressively for each duplicate round.

Final / FSpeed optional

How many initial series clips receive a final speed treatment, and that speed percentage.

Top, Bottom, Left, Right optional

Crop percentages used for the generated first and last crop clips.

Interpolate

Retime a single video or every supported video in a folder. The app reads a sample video and keeps duration, speed, size, and aspect-ratio controls in sync.

Source type and input required

Select File for one video or Folder for a batch. Supported video inputs include MP4, MOV, MKV, WebM, and AVI.

New duration or Speed % one required

Set either target duration or playback speed. Editing one calculates the other. Above 100% is faster; below 100% is slower.

Output FPS optional

Target frame rate. Higher FPS adds frames; lower FPS removes frames.

Width, Height, Resize % optional

Target size controls. These are standard resizes, not AI upscaling. Lock proportions keeps the selected aspect ratio.

Filename suffix optional

Legacy label shown in the UI; output folders are grouped by speed and FPS.

Trim Clip

Inspect a video one frame at a time, drag the green start and purple end handles to select a range, then export the trimmed clip or the currently selected still frame.

• Video is required. Free mode supports videos up to approximately two minutes, with a 2:10 tolerance.
• <<, <, >, >> move to the first frame, previous frame, next frame, and last frame.
• Trim clip saves a new _trimmed video beside the source without overwriting an existing output.
• Export frame saves the current frame as a JPG beside the source.

Image Effect

Preview a folder of images, tune an effect stack, and process the active image or the full folder. Mouse-wheel zoom and drag are available in the preview. Double-clicking an effect slider resets it to its default.

Folder required

Folder containing the images to browse and process.

Saturation optional

From -100 to +100.

Noise, Bloom optional

Each from 0 to 100.

Contrast, Brightness optional

Each from -10 to +10.

Red, Green, Blue optional

Per-channel color adjustment controls.

Presets optional

Save the current values as a named preset, select a saved preset, or delete one.

Previous and Next browse files. Test previews the effect, Show Original toggles the source view, and the two processing buttons write either one output or a full batch.

Cropper

Use one crop rectangle across images in a folder. Drag inside the green rectangle to reposition it; double-click inside it, or press Enter, to crop the current image.

Folder required

Folder containing the images to crop.

Suffix optional

Text appended to each cropped filename; default is _cropped.

Keep proportions optional

Preserves the crop rectangle’s aspect ratio while adjusting it.

File handling optional

Ask, Overwrite, or Extra copy when the target output already exists.

Use arrow keys for previous/next, Home/End for first/last, Reset crop for the full image, and Crop to save.

Encode Brand

Create a branded MP4 from one video. A selected logo can be placed on an opening intro segment and a selected watermark can be placed at lower-right. Outputs are written under a Branded folder beside the source video.

Video required

Source video to brand. Free mode supports videos up to approximately two minutes, with a 2:10 tolerance.

Brand required when branding

Brand kit created in Settings. The built-in 21plus option needs its overlay asset; custom kits can include a logo and/or watermark.

Intro logo optional

Adds the selected logo to an intro segment at the beginning.

Watermark optional

Adds a low-opacity watermark at lower-right when that brand has one.

3D side-by-side optional

Places logo and watermark appropriately in both halves of a side-by-side 3D video.

3D OpenPose Kit

The local 3D editor lets you position a mannequin, adjust bones and hands, save reusable poses, and generate reference maps for ComfyUI or another workflow. It works locally and does not require an online editor service.

• Pose name names saved scene JSON and generated image outputs.
• Save Pose, Saved poses, Load, and Delete manage app-owned pose scenes.
• Output selects a preset. Pose PNG only creates *_pose.png; All maps and Masks/details bundle create pose, depth, normal, and canny PNGs.
• Generate asks the embedded editor to create the selected outputs. When a valid ComfyUI root is configured, files go to its output folder; otherwise they use the app output folder.

Settings

Settings are saved locally for the current Windows installation.

• Ai Render Helper Pro: enter a PayPal transaction ID or payer ID and select Validate Pro. Pro removes the free video-duration and duplicate-pass limits.
• ComfyUI: set the ComfyUI root folder, local API URL, and an optional starter executable or script. Choose the WAN 2.2 model used by Scenario Maker, then use Scan Current Setup or Rescan to validate that selected model, its VAE, text encoder, LoRAs, and custom-node packages at their actual discovered locations. Optional WAN variants are not required.
• Updates and website: Check for Updates compares the installed version with the public release metadata. Open RenderHelper.app opens the website for downloads, purchase, and support.
• Brands: supply a safe brand name plus transparent PNG logo and/or watermark, then choose Create New Brand. The app copies them into its local overlays folder and makes them available in Encode Brand.

Windows automation

Use the command line

The CLI returns structured JSON so PowerShell, a desktop wrapper, or an AI-built interface can read success, errors, output files, warnings, and next actions. It intentionally exposes only the commands listed below; visual-only tools stay in the desktop app.

Set-Location -LiteralPath "C:\Path\To\Ai Render Helper\ARH resources"
$env:PYTHONPATH = ".\src"
python -m ai_render_helper commands schema

General shape:

python -m ai_render_helper <section> <action> --parameter value

Parameter names use --kebab-case on the command line and are converted to snake_case internally. A parameter written without a value becomes true. Quote Windows paths containing spaces.

Supported command reference

For builders

Build your own interface with an AI

Download the Ai Render Helper CLI prompt, put it in your project folder, and give it to your preferred vibe-coding AI. The prompt tells the AI which commands exist, which parameters are required or optional, how to invoke them on Windows, and how to consume JSON results.

1. Download ai-render-helper-cli-prompt.md into your project.
2. Tell your AI: “Read this file and build a local Windows interface that invokes only this documented CLI.”
3. Have the interface call commands schema before relying on a command, validate user inputs, and display the returned JSON result.
4. Use the CLI for supported operations; direct users to the Ai Render Helper desktop window for visual-only tools.

Limits and support

Free mode limits certain video workflows to approximately two minutes and limits Duplicate Speed Up to one duplicate round. Pro validation removes these production limits. Keep original media until you have reviewed the generated result.

For purchase questions, installation help, or a reproducible processing problem, use Contact and include the app version, Windows version, selected tool, source-media details, and relevant log excerpt.