Skip to main content
Glama
kanto-labs

kanto-labs-mcp

Official
by kanto-labs
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript
Local

Extract frames from videos (coming soon)

extract_video_frames
Read-only

Extract still frames from video files via direct URL, producing storyboards, scene cuts, or evenly spaced frames with timestamps and a contact sheet for vision models and thumbnails.

Instructions

COMING SOON: this Kanto Labs actor is not public on the Apify Store yet. Until it is, calls return an 'actor not yet public' error and nothing is charged. Extract still frames from video files reachable by direct URL (MP4, MOV, WEBM, MKV, AVI, M4V, MPEG-TS; not YouTube or TikTok pages): an automatic storyboard of scene changes, scene cuts only, N evenly spaced frames, or one frame every N seconds. Returns public image URLs with timestamps for every frame, a free contact-sheet grid image, and the video's duration, resolution, fps and codecs. Use it to let a vision model 'watch' a video, to make thumbnails or storyboards, or to find scene changes. A call must finish within 300 seconds, so send one or a few short videos per call. Cost: $0.002 per frame extracted (default 12 frames per video = $0.024), billed to your Apify account; videos that fail to download or decode are free. Runs the Apify actor kantolabs/video-frame-extractor with your APIFY_TOKEN; maxTotalChargeUsd caps the spend per call.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
modeNo"auto" (default): first frame + strongest scene changes, topped up with evenly spaced frames to frameCount. "scenes": only real cuts (up to maxFramesPerVideo). "count": exactly frameCount evenly spaced frames. "interval": one frame every intervalSeconds (up to maxFramesPerVideo).
maxWidthNoScale frames down to this width in px (0 = original). Default 1280, plenty for vision models.
videoUrlsYesDirect links to video files, e.g. ["https://media.w3.org/2010/05/sintel/trailer.mp4"].
frameCountNoFrames per video in auto and count modes. Default 12. Each frame is charged.
imageFormatNoFrame image format. Default jpg.
timeoutSecsNoGive up after this many seconds (max 300, Apify's limit for synchronous runs). Default 280. For big batches, split the input into several calls.
contactSheetNoAlso build one grid image of all frames with timecodes (free). Default true.
sceneThresholdNoScene-change sensitivity (lower finds subtler changes). Default "0.3".
intervalSecondsNoSeconds between frames in interval mode. Default 10.
maxFramesPerVideoNoUpper limit per video in scenes and interval modes. Default 50.
maxTotalChargeUsdNoHard spending cap for this call in USD, enforced by Apify: the run stops once it has charged this much. Default 1 (server setting KANTO_MAX_CHARGE_USD).

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv1.2.0

TDQS

A4.7/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already mark readOnly and non-destructive; the description adds cost per frame, free failed videos, an error when the actor isn't public, and the 300-second timeout. These go beyond annotation fields and give accurate expectations for invocation.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Long but every sentence carries value: the 'coming soon' warning is front-loaded, followed by supported formats, modes, outputs, use cases, constraints, and cost. No filler or repetition.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Covers input constraints (direct URLs, formats, exclusions), output (image URLs, timestamps, contact sheet, metadata), cost model, error behavior, and timeout. With 11 parameters and no output schema, this is comprehensive enough for an agent to call correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the schema already documents every parameter. The description mentions the default frame count and cost, but doesn't add syntax or format beyond what the schema provides; the 'not YouTube/TikTok' note is helpful but already implied by 'direct URL'. Baseline 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States exactly what it does: extracts still frames from video files via direct URL, lists supported formats and exclusions (YouTube/TikTok), enumerates output (image URLs, timestamps, contact sheet, metadata), and gives use cases. It clearly distinguishes from sibling tools, which are unrelated to video.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly provides use cases ('let a vision model watch a video, make thumbnails or storyboards, find scene changes') and exclusions ('not YouTube or TikTok pages'). It also advises on batch size due to the 300-second limit, which helps an agent decide when to call.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools