根据文本提示生成视频

REST

如需详细了解如何使用 Gemini Omni Flash API，请参阅Interactions API。

生成视频可能需要超过一分钟才能完成。如需生成视频并在完成后立即下载，请使用同步请求。如需生成可供日后下载的视频，请将 background 参数设置为 true，以发送异步请求。异步互动内容最多可保留 14 天。

POST https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions

{
  "model": "MODEL_ID",
  "input": [
    {
      "type": "text",
      "text": "TEXT_PROMPT"
    }
  ],
  "response_format": [
    {
      "type": "video",
      "delivery": "uri",
      "gcs_uri": "CLOUD_STORAGE_OUTPUT_URI",
      "aspect_ratio": "ASPECT_RATIO",
      "duration": "DURATION"
    }
  ],
  "generation_config": {
    "video_config": {
      "task": "text_to_video"
    }
  }
}

curl -X POST \
     -H "Authorization: Bearer TOKEN" \
     -H "Content-Type: application/json; charset=utf-8" \
     -d @request.json \
     "https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions"

$headers = @{ "Authorization" = "Bearer TOKEN" }

Invoke-WebRequest `
    -Method POST `
    -Headers $headers `
    -ContentType: "application/json; charset=utf-8" `
    -InFile request.json `
    -Uri "https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions" | Select-Object -Expand Content

{
  "id":"INTERACTION_ID",
  "model":"gemini-omni-flash-preview",
  "status":"completed",
  "usage":{
    "total_tokens":479,
    "total_input_tokens":26,
    "input_tokens_by_modality":[
      {
        "modality":"text",
        "tokens":26
      },
      {
        "modality":"image",
        "tokens":124
      }
    ],
    "output_tokens_by_modality": [
      {
        "modality": "video",
        "tokens": 28832
      }
    ],
    "total_output_tokens":28832,
    "total_thought_tokens":453
  },
  "steps":[
    {
      "type":"thought",
      "summary":[
        {
          "type":"text",
          "text":"MODEL THOUGHTS"
        }
      ],
    },
    { 
      "type":"model_output",
      "content":[
        {
          "type":"video",
          "uri":"gs://some/output_path/123.mp4",
          "mime_type":"video/mp4" 
        }
      ]
    }
  ],
  "object":"interaction"
  "role":"model",
  "created":"2026-05-29T02:17:56Z",
  "updated":"2026-05-29T02:17:56Z",
}

POST https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions

{
  "model": "MODEL_ID",
  "input": [
    {
      "background": true,
      "type": "text",
      "text": "TEXT_PROMPT"
    }
  ],
  "response_format": [
    {
      "type": "video",
      "delivery": "uri",
      "gcs_uri": "CLOUD_STORAGE_OUTPUT_URI",
      "aspect_ratio": "ASPECT_RATIO",
      "duration": "DURATION"
    }
  ],
  "generation_config": {
    "video_config": {
      "task": "text_to_video"
    }
  }
}

curl -X POST \
     -H "Authorization: Bearer TOKEN" \
     -H "Content-Type: application/json; charset=utf-8" \
     -d @request.json \
     "https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions"

$headers = @{ "Authorization" = "Bearer TOKEN" }

Invoke-WebRequest `
    -Method POST `
    -Headers $headers `
    -ContentType: "application/json; charset=utf-8" `
    -InFile request.json `
    -Uri "https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions" | Select-Object -Expand Content

{
  "id":"INTERACTION_ID",
  "status":"in_progress",
  "object":"interaction"
}

POST https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions/INTERACTION_ID

curl -X POST \
     -H "Authorization: Bearer TOKEN" \
     -H "Content-Type: application/json; charset=utf-8" \
     -d "" \
     "https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions/INTERACTION_ID"

$headers = @{ "Authorization" = "Bearer TOKEN" }

Invoke-WebRequest `
    -Method POST `
    -Headers $headers `
    -Uri "https://aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/global/interactions/INTERACTION_ID" | Select-Object -Expand Content

{
  "id":"INTERACTION_ID",
  "model":"gemini-omni-flash-preview",
  "status":"completed",
  "usage":{
    "total_tokens":479,
    "total_input_tokens":26,
    "input_tokens_by_modality":[
      {
        "modality":"text",
        "tokens":26
      },
      {
        "modality":"image",
        "tokens":124
      }
    ],
    "output_tokens_by_modality": [
      {
        "modality": "video",
        "tokens": 28832
      }
    ],
    "total_output_tokens":28832,
    "total_thought_tokens":453
  },
  "steps":[
    {
      "type": "user_input",
      "content": [
        {
          "type": "text",
          "text": "5 second, 9:16 video. Use the image as the first frame."
        },
        {
          "type": "image",
          "uri": "gs://some/path",
          "mime_type": "image/png"
        }
      ]
    },
    {
      "type":"thought"
      "summary":[
        {
          "type":"text",
          "text":"MODEL THOUGHTS"
        }
      ],
    },
    { 
      "type":"model_output",
      "content":[
        {
          "type":"video",
          "data":"VIDEO DATA",
          "mime_type":"video/mp4" 
        }
      ]
    }
  ],
  "object":"interaction"
  "role":"model",
  "created":"2026-05-29T02:17:56Z",
  "updated":"2026-05-29T02:17:56Z",
}

控制台

1. 在 Google Cloud 控制台中，依次前往 Agent Platform > Media Studio 页面。

   Media Studio

2. 点击视频。

3. 在任务菜单中，选择文字转视频。

4. 在模型菜单中，从显示的选项中选择一个模型。

5. 在提示框中，输入描述要生成的视频的文本提示。

6. 可选：调整以下参数：

   • 宽高比：选择 16:9 或 9:16。

   • 结果数量：调整滑块或输入介于 1 和 4 之间的值。

   • 视频时长：从菜单中选择视频时长。

   • 输出目录：点击浏览以创建或选择一个 Cloud Storage 存储桶来存储生成的文件。

7. 点击运行。

Python

安装

pip install --upgrade google-genai

如需了解详情，请参阅 SDK 参考文档。

设置环境变量以将 Google Gen AI SDK 与 Vertex AI 搭配使用：

# Replace the `GOOGLE_CLOUD_PROJECT` and `GOOGLE_CLOUD_LOCATION` values
# with appropriate values for your project.
export GOOGLE_CLOUD_PROJECT=GOOGLE_CLOUD_PROJECT
export GOOGLE_CLOUD_LOCATION=global
export GOOGLE_GENAI_USE_ENTERPRISE=True

import time
from google import genai
from google.genai.types import GenerateVideosConfig

client = genai.Client()

# TODO(developer): Update and un-comment below line
# output_gcs_uri = "gs://your-bucket/your-prefix"

operation = client.models.generate_videos(
    model="veo-3.1-generate-001",
    prompt="a cat reading a book",
    config=GenerateVideosConfig(
        aspect_ratio="16:9",
        output_gcs_uri=output_gcs_uri,
    ),
)

while not operation.done:
    time.sleep(15)
    operation = client.operations.get(operation)
    print(operation)

if operation.response:
    print(operation.result.generated_videos[0].video.uri)

# Example response:
# gs://your-bucket/your-prefix

Go

了解如何安装或更新 Go。

如需了解详情，请参阅 SDK 参考文档。

设置环境变量以将 Google Gen AI SDK 与 Vertex AI 搭配使用：

# Replace the `GOOGLE_CLOUD_PROJECT` and `GOOGLE_CLOUD_LOCATION` values
# with appropriate values for your project.
export GOOGLE_CLOUD_PROJECT=GOOGLE_CLOUD_PROJECT
export GOOGLE_CLOUD_LOCATION=global
export GOOGLE_GENAI_USE_ENTERPRISE=True

import (
	"context"
	"fmt"
	"io"
	"time"

	"google.golang.org/genai"
)

// generateVideoWithText shows how to gen video from text.
func generateVideoWithText(w io.Writer, outputGCSURI string) error {
	//outputGCSURI = "gs://your-bucket/your-prefix"
	ctx := context.Background()

	client, err := genai.NewClient(ctx, &genai.ClientConfig{
		HTTPOptions: genai.HTTPOptions{APIVersion: "v1"},
	})
	if err != nil {
		return fmt.Errorf("failed to create genai client: %w", err)
	}

	config := &genai.GenerateVideosConfig{
		AspectRatio:  "16:9",
		OutputGCSURI: outputGCSURI,
	}

	modelName := "veo-3.0-generate-preview"
	prompt := "a cat reading a book"
	operation, err := client.Models.GenerateVideos(ctx, modelName, prompt, nil, config)
	if err != nil {
		return fmt.Errorf("failed to start video generation: %w", err)
	}

	// Polling until the operation is done
	for !operation.Done {
		time.Sleep(15 * time.Second)
		operation, err = client.Operations.GetVideosOperation(ctx, operation, nil)
		if err != nil {
			return fmt.Errorf("failed to get operation status: %w", err)
		}
	}

	if operation.Response != nil && len(operation.Response.GeneratedVideos) > 0 {
		videoURI := operation.Response.GeneratedVideos[0].Video.URI
		fmt.Fprintln(w, videoURI)
		return nil
	}

	// Example response:
	// gs://your-bucket/your-prefix/videoURI

	return fmt.Errorf("video generation failed or returned no results")
}

Node.js

安装

npm install @google/genai

如需了解详情，请参阅 SDK 参考文档。

设置环境变量以将 Google Gen AI SDK 与 Vertex AI 搭配使用：

# Replace the `GOOGLE_CLOUD_PROJECT` and `GOOGLE_CLOUD_LOCATION` values
# with appropriate values for your project.
export GOOGLE_CLOUD_PROJECT=GOOGLE_CLOUD_PROJECT
export GOOGLE_CLOUD_LOCATION=global
export GOOGLE_GENAI_USE_ENTERPRISE=True

const {GoogleGenAI} = require('@google/genai');

const GOOGLE_CLOUD_PROJECT = process.env.GOOGLE_CLOUD_PROJECT;
const GOOGLE_CLOUD_LOCATION = process.env.GOOGLE_CLOUD_LOCATION || 'global';

async function generateVideo(
  outputGcsUri,
  projectId = GOOGLE_CLOUD_PROJECT,
  location = GOOGLE_CLOUD_LOCATION
) {
  const client = new GoogleGenAI({
    vertexai: true,
    project: projectId,
    location: location,
  });

  let operation = await client.models.generateVideos({
    model: 'veo-3.1-fast-generate-001',
    prompt: 'a cat reading a book',
    config: {
      aspectRatio: '16:9',
      outputGcsUri: outputGcsUri,
    },
  });

  while (!operation.done) {
    await new Promise(resolve => setTimeout(resolve, 15000));
    operation = await client.operations.get({operation: operation});
    console.log(operation);
  }

  if (operation.response) {
    console.log(operation.response.generatedVideos[0].video.uri);
  }
  return operation;
}

Java

了解如何安装或更新 Java。

如需了解详情，请参阅 SDK 参考文档。

设置环境变量以将 Google Gen AI SDK 与 Vertex AI 搭配使用：

# Replace the `GOOGLE_CLOUD_PROJECT` and `GOOGLE_CLOUD_LOCATION` values
# with appropriate values for your project.
export GOOGLE_CLOUD_PROJECT=GOOGLE_CLOUD_PROJECT
export GOOGLE_CLOUD_LOCATION=global
export GOOGLE_GENAI_USE_ENTERPRISE=True

import com.google.genai.Client;
import com.google.genai.types.GenerateVideosConfig;
import com.google.genai.types.GenerateVideosOperation;
import com.google.genai.types.GenerateVideosResponse;
import com.google.genai.types.GenerateVideosSource;
import com.google.genai.types.GeneratedVideo;
import com.google.genai.types.GetOperationConfig;
import com.google.genai.types.Video;
import java.util.concurrent.TimeUnit;

public class VideoGenWithTxt {

  public static void main(String[] args) throws InterruptedException {
    // TODO(developer): Replace these variables before running the sample.
    String modelId = "veo-3.1-generate-001";
    String outputGcsUri = "gs://your-bucket/your-prefix";
    generateContent(modelId, outputGcsUri);
  }

  // Generates a video with a text prompt.
  public static String generateContent(String modelId, String outputGcsUri)
      throws InterruptedException {
    // Client Initialization. Once created, it can be reused for multiple requests.
    try (Client client = Client.builder().location("global").vertexAI(true).build()) {

      GenerateVideosOperation operation =
          client.models.generateVideos(
              modelId,
              GenerateVideosSource.builder().prompt("a cat reading a book").build(),
              GenerateVideosConfig.builder()
                  .aspectRatio("16:9")
                  .outputGcsUri(outputGcsUri)
                  .build());

      while (!operation.done().orElse(false)) {
        TimeUnit.SECONDS.sleep(15);
        operation =
            client.operations.getVideosOperation(operation, GetOperationConfig.builder().build());
      }

      String generatedVideoUri =
          operation
              .response()
              .flatMap(GenerateVideosResponse::generatedVideos)
              .flatMap(videos -> videos.stream().findFirst())
              .flatMap(GeneratedVideo::video)
              .flatMap(Video::uri)
              .orElseThrow(
                  () ->
                      new IllegalStateException(
                          "Could not get the URI from the generated video"));

      System.out.println("Generated video URI: " + generatedVideoUri);
      // Example response:
      // Generated video URI: gs://your-bucket/your-prefix/generated-video-123.mp4
      return generatedVideoUri;
    }
  }
}

REST

如需详细了解 Veo API，请参阅以下内容：

• 方法： endpoints.predict
• VideoGenerationModelInstance
• VideoGenerationModelParams
• VideoGenerationModelResult

1. 使用以下命令发送视频生成请求。此请求会启动长时间运行的操作，并将输出存储到您指定的 Cloud Storage 存储桶。

   在使用任何请求数据之前，请先进行以下替换：

   • PROJECT_ID：表示您的 Google Cloud 项目 ID 的字符串。
   • MODEL_ID：表示要使用的模型 ID 的字符串。接受的值如下：
     • Veo 2： "veo-2.0-generate-001"
     • Veo 3："veo-3.0-generate-001"
     • Veo 3："veo-3.0-fast-generate-001"
     • Veo 3："veo-3.0-generate-preview"（预览版）
     • Veo 3："veo-3.0-fast-generate-preview"（预览版）
     • Veo 3.1： "veo-3.1-generate-001"
     • Veo 3.1： "veo-3.1-fast-generate-001"
   • TEXT_PROMPT：用于指导视频生成的文本提示。
   • OUTPUT_STORAGE_URI：（可选）一个字符串，表示用于存储输出视频的 Cloud Storage 存储桶。如果未提供，则回答中会返回视频字节。例如："gs://video-bucket/output/"
   • RESPONSE_COUNT：要生成的视频文件数量。可接受的值范围为 1-4。
   • DURATION：一个整数，表示生成的视频文件的时长。以下是每种模型接受的值：
     • Veo 2 模型：5-8。默认值为 8。
     • Veo 3 模型：4、6 或 8。默认为 8。

   • 其他可选参数

     根据您的应用场景，使用以下可选变量。在 "parameters": {} 对象中添加以下部分或全部参数。

     "parameters": {
       "aspectRatio": "ASPECT_RATIO",
       "negativePrompt": "NEGATIVE_PROMPT",
       "personGeneration": "PERSON_SAFETY_SETTING",
       // "resolution": RESOLUTION, // Veo 3 models only
       "sampleCount": RESPONSE_COUNT,
       "seed": SEED_NUMBER
     }

     • ASPECT_RATIO：（可选）一个字符串值，用于描述所生成视频的宽高比。您可使用以下值：
       • "16:9" 表示横向
       • "9:16" 表示纵向

       默认值为 "16:9"

     • NEGATIVE_PROMPT：（可选）一个字符串值，用于描述您想要阻止模型生成的内容。
     • PERSON_SAFETY_SETTING：（可选）一个字符串值，用于控制生成人物或人脸的安全设置。您可使用以下值：
       • "allow_adult"：仅允许生成成人形象和人脸。
       • "disallow"：不生成人物或人脸。

       默认值为 "allow_adult"。

     • RESOLUTION：（可选）用于控制生成的视频分辨率的字符串值。仅受 Veo 3 模型支持。您可使用以下值：
       • "720p"
       • "1080p"
       • "4k"（仅限 Veo 3.1 预览版模型）

       默认值为 "720p"。

     • RESPONSE_COUNT：可选。一个整数值，用于描述要生成的视频数量。可接受的值范围为 1-4。
     • SEED_NUMBER：可选。一个 uint32 值，模型使用该值生成确定性视频。通过在请求中指定种子编号而无需更改其他参数，可引导模型生成相同的视频。可接受的值范围为 0-4294967295。

   HTTP 方法和网址：

   POST https://us-central1-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID:predictLongRunning

   请求 JSON 正文：

   {
     "instances": [
       {
         "prompt": "TEXT_PROMPT"
       }
     ],
     "parameters": {
       "storageUri": "OUTPUT_STORAGE_URI",
       "sampleCount": "RESPONSE_COUNT"
     }
   }
   

   如需发送请求，请选择以下方式之一：

   curl

   将请求正文保存在名为 request.json 的文件中，然后执行以下命令：

   curl -X POST \
        -H "Authorization: Bearer $(gcloud auth print-access-token)" \
        -H "Content-Type: application/json; charset=utf-8" \
        -d @request.json \
        "https://us-central1-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID:predictLongRunning"

   PowerShell

   将请求正文保存在名为 request.json 的文件中，然后执行以下命令：

   $cred = gcloud auth print-access-token
   $headers = @{ "Authorization" = "Bearer $cred" }
   
   Invoke-WebRequest `
       -Method POST `
       -Headers $headers `
       -ContentType: "application/json; charset=utf-8" `
       -InFile request.json `
       -Uri "https://us-central1-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID:predictLongRunning" | Select-Object -Expand Content

   此请求会返回包含唯一操作 ID 的完整操作名称。使用此完整操作名称轮询视频生成请求的状态。
   {
     "name": "projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID/operations/a1b07c8e-7b5a-4aba-bb34-3e1ccb8afcc8"
   }
   

2. 可选：检查长时间运行的视频生成操作的状态。

   在使用任何请求数据之前，请先进行以下替换：

   • PROJECT_ID：您的 Google Cloud 项目 ID。
   • MODEL_ID：要使用的模型 ID。
   • OPERATION_ID：在原始生成视频请求中返回的唯一操作 ID。

   HTTP 方法和网址：

   POST https://us-central1-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID:fetchPredictOperation

   请求 JSON 正文：

   {
     "operationName": "projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID/operations/OPERATION_ID"
   }
   

   如需发送请求，请选择以下方式之一：

   curl

   将请求正文保存在名为 request.json 的文件中，然后执行以下命令：

   curl -X POST \
        -H "Authorization: Bearer $(gcloud auth print-access-token)" \
        -H "Content-Type: application/json; charset=utf-8" \
        -d @request.json \
        "https://us-central1-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID:fetchPredictOperation"

   PowerShell

   将请求正文保存在名为 request.json 的文件中，然后执行以下命令：

   $cred = gcloud auth print-access-token
   $headers = @{ "Authorization" = "Bearer $cred" }
   
   Invoke-WebRequest `
       -Method POST `
       -Headers $headers `
       -ContentType: "application/json; charset=utf-8" `
       -InFile request.json `
       -Uri "https://us-central1-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID:fetchPredictOperation" | Select-Object -Expand Content

   此请求会返回有关操作的信息，包括操作是否仍在运行或已完成。

   响应

   {
     "name": "projects/PROJECT_ID/locations/us-central1/publishers/google/models/MODEL_ID/operations/OPERATION_ID",
     "done": true,
     "response": {
       "raiMediaFilteredCount": 0,
       "@type": "type.googleapis.com/cloud.ai.large_models.vision.GenerateVideoResponse",
       "videos": [
         {
           "gcsUri":"gs://BUCKET_NAME/TIMESTAMPED_FOLDER/sample_0.mp4",
           "mimeType": "video/mp4"
         }
       ]
     }
   }