<aside> 🛠

Pull a Single Frame Out of a Video as an Image

Pick any frame from an input video by its 0-based index and pop it out as a standalone image. Perfect for bridging video and image pipelines: Extract a key frame, run it through image super-resolution (PiSA-SR), an image generator (Nano-Banana), or any image-only tool, then feed the result back into a video super-resolution model like SparkVSR.

</aside>

What it does

Video Frame Extract reads a video and returns one specific frame as a still image. You set the 0-based frame index (0 = first frame, 1 = second frame, and so on) and the tool decodes only that frame and emits it as a standard image — ready to feed into any image-processing node in the studio. The video itself isn't modified; only the chosen frame is extracted. The operation is CPU-only and runs instantly.

Problem it solves

• Bridge video → image workflows – Feed a video frame into any image-only tool (upscalers, generators, color tools, masking tools)
• Pre-process key frames – Extract a reference frame, enhance it with image models (e.g. PiSA-SR, Nano-Banana), then feed back into a video super-resolution model (e.g. SparkVSR)
• Thumbnails & previews – Pull a representative frame for posters, thumbnails, or social cards
• Frame-by-frame inspection – Extract a specific frame for visual QA, debugging, or annotation
• Reference-image conditioning – Use a video frame as a conditioning image for image generation / edit models
• Quick "before" snapshot – Capture the source video's frame 0 before running video enhancement, for side-by-side comparison with Image Compare

Input/Output

<aside>

• Input: Single video

  • Format: Standard video formats (MP4, MOV, WebM, etc.)
  • Length / resolution: Any (the tool only decodes the requested frame)
  • Source: Any upstream video node or uploaded video file

  ebf8274b-d767-4c55-8bae-10e545d3fe16.mp4

</aside>

<aside>

• Output: Single image (the extracted frame)
  • Format: Standard image format (PNG / JPEG)
  • Resolution: Matches the video's native frame resolution
  • Content: Exactly the frame at the requested index — no scaling, no re-encoding artifacts beyond standard decode </aside>

Configuration Options

• Frame index (0-based): Which frame to pull out.
  • Range: 0 to (total frames - 1)
  • 0 - First frame of the video (often the "establishing" frame, good for thumbnails / posters)
  • N - The (N+1)th frame
  • Behavior: The tool decodes only the requested frame, so requesting a late frame in a long video is fast - no full pass through the video needed
  • Tip: If you want a frame at a specific timestamp, compute the index as time_in_seconds × frame_rate

How to Use

<aside>

1. Load video → Connect an upstream video source (uploaded video, video generation node, etc.) to Video Frame Extract
2. Choose a frame → Enter the 0-based frame index you want to extract (0 for the very first frame)
3. Run Flow → The selected frame is decoded and emitted as an image, ready to feed into any downstream image tool

</aside>

Use Case Tips

For Video → Image Super-Resolution → Video Super-Resolution Pipeline: