Extract Frame

Grab one still from a video (first, last, middle, or a custom timestamp) as COMFYTV_IMAGE. Uses PyAV on disk — no GPU.

Extract Frame · screenshot coming soon

What this node does

Extract Frame freezes one frame from a COMFYTV_VIDEO clip. Pick position or at_seconds, get an image snapshot. Common for ↪ Extend (last frame → new I2V), thumbnails, or Image Edit / Upscale.

Processing uses PyAV (FFmpeg bindings) on the file behind the /view? URL — fast, zero VRAM.

When to use it

• Last frame from an AI clip for I2V continuation
• Inspect motion or composition at a specific time
• Video → still → edit / upscale → back to video

How ComfyTV designed this

• Stage + ▶ Run: extracts from upstream video snapshot only.
• Snapshot: reads upstream video URL; re-run upstream after changes.
• PyAV backend: decode one frame, write PNG. See video-and-audio.md.

Types (COMFYTV_* vs native ComfyUI)

ComfyTV type	What it is	vs ComfyUI
COMFYTV_VIDEO	Video URL snapshot	Not a ComfyUI VIDEO object
COMFYTV_IMAGE	Extracted frame URL	Not an IMAGE tensor

Conversion: plugin video → → ComfyTV Video; frame to tensor → ← ComfyTV Image. See bridges.md.

Parameters

video (input)

Upstream COMFYTV_VIDEO. Unwired → Run error.

position

• last — final frame (extend workflows)
• first — opening frame
• middle — midpoint
• custom — uses at_seconds

at_seconds

Timestamp in seconds when position=custom. Range 0–3600.

Outputs

Output	Type	Meaning	Downstream
image	COMFYTV_IMAGE	Extracted frame	Upscale, Image Edit, Video Stage (I2V)

Step by step

1. Upstream node outputs COMFYTV_VIDEO (e.g. Video Stage), Run it.
2. Add Extract Frame, wire video.
3. Set position (e.g. last) or custom + seconds.
4. ▶ Run, check thumbnail.
5. Wire Upscale or Video Stage (I2V).

Full guides (recommended reading)

This page covers one node only. For end-to-end workflows, multi-stage pipelines, type conversion, and design rationale, see the ComfyTV user guides on GitHub:

Guide	Contents
Getting started	Install, canvas basics, per-stage Run, snapshots, Project, Image Picker
Video and audio	Clip, crop, resize, extract frame, demux vs Generate video

Repository and workflows

Resource	Link
GitHub repository	https://github.com/jtydhr88/ComfyTV
User guides index	https://github.com/jtydhr88/ComfyTV/tree/main/docs
Built-in workflows	https://github.com/jtydhr88/ComfyTV/tree/main/workflows
Model checklist	https://github.com/jtydhr88/ComfyTV/blob/main/docs/models.md
Custom workflows	https://github.com/jtydhr88/ComfyTV/blob/main/docs/custom-workflows.md

FAQ

Q: “needs upstream video”?
A: Run upstream video stage; confirm video is wired.

Q: vs Load Video + Save Image?
A: Extract Frame stays in ComfyTV snapshots (COMFYTV_IMAGE downstream) and runs this node only.

Q: Custom time beyond duration?
A: PyAV clamps to valid range; may land near the last frame.

Related nodes

• Video Stage / Load Video
• Video Clip
• ↪ Extend (toolbar) — auto last frame + new Video Stage