Google Vertex AI icon

Google Vertex AI

Use Google Vertex AI (Gemini) for text, media analysis, and image generation/editing

Actions11

Video → Analyze Video

AI-generated

Summary

Sends a video plus a text prompt to a Gemini model on Vertex AI (generateContent) and returns the model's answer about the video, as plain text by default.

Inputs

  • Input Type — Where the video comes from: 'Binary File' (default) from the incoming item, or 'Google Cloud Storage URI'.
  • Input Binary Field — Binary property holding the video (default 'data'). Used when Input Type is Binary File. The file is read and sent inline as base64.
  • GCS URI — gs://bucket/path/file location of the video. Used when Input Type is Google Cloud Storage URI.
  • MIME Type — e.g. video/mp4. Optional for binary (falls back to the binary's detected type), but required for a GCS URI. The node throws an error if it is missing.
  • Prompt (required) — The question or instruction about the video.
  • Model (required) — Gemini model, picked from a list or entered by ID (default gemini-2.5-flash). It is read by the source but not in the filtered property list.
  • Options — Optional collection. Includes System Instruction, Temperature, Top K, Top P, Max Output Tokens, JSON Output (requests application/json), and Simplify Output (default on). These are read by the source but not in the filtered property list.

Output shape

One output item per input item. With Simplify Output on (default), the JSON is { text }, the concatenated text of the first candidate. With it off, the JSON is the full Vertex AI generateContent response.

Large binary videos are sent inline as base64, so request size limits may apply; a GCS URI avoids this. Authentication uses a service-account credential, and the project and region come from that credential. With Continue On Fail, errors are returned as { error } items. API errors for 401/403/404/429 get explanatory hints.

Examples

Example 1: Summarize an uploaded MP4 from a previous node

Input Type: Binary File; Input Binary Field: data; Prompt: 'Summarize this video in 5 bullet points'; Model: gemini-2.5-flash.

Example 2: Analyze a video stored in Cloud Storage

Input Type: Google Cloud Storage URI; GCS URI: gs://my-bucket/clips/demo.mp4; MIME Type: video/mp4; Prompt: 'List the on-screen text'.

Example 3: Extract structured data from a video

Binary input; Prompt describing the fields to extract; Options: JSON Output on, Temperature 0.

Links

Discussion