gemini-video-input

Describe a Youtube Video

See Live Demo built using these code snippets

Javascript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI(apiKey);
const model = genAI.getGenerativeModel({ model: "gemini-2.0-flash" });
const result = await model.generateContent([
  {
    fileData: {
      mimeType: "video/webm",
      fileUri: youtubeUrl
    }
  },
  "Describe what you see in this video in detail."
]);

Python

from google import genai
from google.genai.types import Part

client = genai.Client()
YOUTUBE_VIDEO_URL = (
    "https://www.youtube.com/watch?v=O_W_VGUeHVI"
)
response = client.models.generate_content(
    model="gemini-2.0-flash",
    contents=[
        Part.from_uri(
            file_uri=YOUTUBE_VIDEO_URL,
            mime_type="video/webm",
        ),
        "Describe what you see in this video in detail."
    ]
)

print(response.text)

Describing local videos with Gemini

See Live Demo built using these code snippets

Javascript

import { GoogleGenerativeAI } from "@google/generative-ai";
import * as fs from "node:fs";

async function describeVideo(videoPath, apiKey) {
  // Initialize the Gemini API client
  const genAI = new GoogleGenerativeAI(apiKey);
  const model = genAI.getGenerativeModel({ model: "gemini-2.0-flash" });

  try {
    // Read the video file and convert to base64
    const videoBytes = fs.readFileSync(videoPath, { encoding: "base64" });

    console.log("Generating video description...");
    const result = await model.generateContent([
      "Please describe what you see in this video in detail. Include any notable events, " +
      "people, objects, or actions that appear.",
      {
        inlineData: {
          data: videoBytes,
          mimeType: "video/mp4",
        },
      },
    ]);

    console.log(result.response.text());
  } catch (error) {
    console.error("Error processing video:", error);
    throw error;
  }
}

Python

from google import genai

def describe_video(video_path, api_key):
    client = genai.Client(api_key=api_key)
    
    print("Uploading video file...")
    video_file = client.files.upload(file=video_path)
    print(f"Completed upload: {video_file.uri}")
    
    print("Waiting for video processing...")
    while video_file.state.name == "PROCESSING":
        print('.', end='', flush=True)
        time.sleep(1)
        video_file = client.files.get(name=video_file.name)
    
    if video_file.state.name == "FAILED":
        raise ValueError(f"Video processing failed: {video_file.state.name}")
    
    print("\nGenerating video description...")
    response = client.models.generate_content(
        model="gemini-2.0-flash",
        contents=[
            video_file,
            "Please describe what you see in this video in detail. Include any notable events, "
            "people, objects, or actions that appear."
        ]
    )
    print(response.text)