How do I generate a video using grok-imagine-video API?

You send a POST request with model 'grok-imagine-video' including a text prompt and optional image/video source; it returns a task ID, then poll this ID until the video status is 'done'.

What input formats does grok-imagine-video accept?

It accepts a natural language prompt and optional image URLs (or base64 images) for animation; for editing, a video URL is provided.

What are the maximum video duration and resolution options?

The model supports video generation up to 15 seconds and resolutions up to 720p with configurable aspect ratios like 16:9 or 1:1.

Can I animate a still image into a video with this model?

Yes, you can animate a still image into motion based on your prompt, using image URLs or encoded images in the request.

Does grok-imagine-video support editing existing videos?

Yes — provide the source video URL and your edit instructions; the output keeps the original video’s duration and resolution.

How does moderation work in grok-imagine-video outputs?

Generated videos are subject to content moderation; flagged content may be filtered or blocked during generation.

Is the video generation asynchronous?

Yes, the API returns a request ID which you poll to check when the video is ready for download.

How should I handle expired video URLs?

Downloaded videos should be saved quickly; temporary URLs may expire and become inaccessible after generation.

Affordable Grok Imagine Video API | image-to-video

📘 Technical Specifications of Grok Imagine Video

Specification	Details
Model ID	grok-imagine-video
Provider	xAI
Type	Video generation & editing AI
Input Types	Text (prompt); optional image or video Text prompts (natural language); optional image input (image→video); optional video_url for editing existing clips. Editing input video max durations differ by endpoint — reported ~8.7s for some editing flows.
Output Types	.mp4 video via temporary URL
Duration Range (generate)	1–15 seconds
Resolution	480p, 720p (configurable)
Aspect Ratios	1:1, 16:9, 9:16
Edit Support	Yes — animates & modifies videos up to 8.7s
Moderation	Content moderation included
Pricing	Charged per second, varies by resolution

🚀 What is Grok Imagine Video?

Grok Imagine Video is xAI’s advanced video generation and editing AI model exposed through CometAPI. It lets developers generate short, custom videos from natural language prompts and optionally animate still images or edit existing clips. The model supports configurable output length, resolution, and aspect ratio, with built-in content moderation to ensure policy compliance.

🧠Main features (what differentiates Grok Imagine)

Native audio + lip-sync: Generates synchronized ambient audio, effects, and short speech / narration with approximate lip synchronization.
Image→Video / prompt editing: Animate a still or edit existing footage via text prompts (remove/replace objects, retime, restyle).
Fast iteration & low latency: Designed for quick feedback loops suitable for creative workflows and product prototyping.
Production API: Imagine API exposes programmatic endpoints for batch generation, integration into editing pipelines, and enterprise controls.
Multiple “modes” / styles: User-facing modes (reported examples: Normal / Fun / Spicy or similar presets) to bias outputs for style or permissiveness (note: “Spicy” mode historically enabled NSFW).

Model (company)	Max res (public)	Max clip len (public)	Native audio?	Strengths	Caveats
Grok Imagine (xAI)	720p	6–15s	Yes	Fast iteration, strong cost/latency, integrated editing, native audio	720p cap; moderation concerns; varying real-world fidelity
Sora (OpenAI)	720p–1080p (depends on tier)	short (6–15s)	Yes	High visual fidelity; strong integration with OpenAI stack	More expensive; constrained moderation/controls
Veo (Google DeepMind)	Up to 1080p+	short (varies)	Yes	Strong photorealism, stable motion	Higher cost; less public experimentation
Runway Gen-4.5	1080p+	short (varies)	Yes	Industry adoption for creative workflows, high fidelity	Costlier; focused on creative tooling
Vidu / Kling / Pika (various specialists)	up to 1080p	short (varies)	Mixed	Some offer niche features (Smart Cuts, multi-shot chaining)	Varied audio support; differing API maturity

⚠️ Limitations

Maximum video length is capped at 15 seconds.
Editing retains input video length (≤ 8.7s).
Generated URLs are ephemeral — download promptly.

How to access and integrate Grok Imagine Video

Log in to cometapi.com. If you are not our user yet, please register first. Sign into your CometAPI console. Get the access credential API key of the interface. Click “Add Token” at the API token in the personal center, get the token key: sk-xxxxx and submit.

Step 2: Send Requests to `Grok Imagine Video` API

Select the “grok-imagine-video” endpoint to send the API request and set the request body. The request method and request body are obtained from our website API doc. Our website also provides Apifox test for your convenience. Replace <YOUR_API_KEY> with your actual CometAPI key from your account. Where to call it: GROKVideo Generation and Video Edit.

Step 3: Send Requests to `Grok Imagine Video` API

Enter text or upload an image(You can optionally provide a source image to animate.). The Grok Imagine AI API analyzes your input and prepares the content for url. Both text-to-video and image-to-video conversion are supported.

The source image can be provided as:

A public URL pointing to an image
A base64-encoded data URI( e.g., data:image/jpeg;base64,<YOUR_BASE64_IMAGE>)

Step 4: Retrieve and Verify Results

Process the API response to get the generated answer. After processing, the API responds with the task status and output data. It returns a request_id immediately upon submission; use the GET endpoint to check status and retrieve the generated video. Video editing is asynchronous, you may need to poll this endpoint multiple times until the task is complete. Please download promptly.

Pricing for Grok Imagine Video

Explore competitive pricing for Grok Imagine Video, designed to fit various budgets and usage needs. Our flexible plans ensure you only pay for what you use, making it easy to scale as your requirements grow. Discover how Grok Imagine Video can enhance your projects while keeping costs manageable.

Pricing Overview

Category	Item	Price
Input Pricing	Text	N/A (Free)
	Image	$0.0016
	Video per second	$0.008
Output Pricing	480p	$0.04
(Per second by resolution)	720p	$0.056

Note: When generating video via API, you are charged per second. You will also be charged when using video or images as input.

Sample code and API for Grok Imagine Video

Access comprehensive sample code and API resources for Grok Imagine Video to streamline your integration process. Our detailed documentation provides step-by-step guidance, helping you leverage the full potential of Grok Imagine Video in your projects.

Python
JavaScript
Curl

import os
import time
import requests

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com/grok/v1"

headers = {
    "Authorization": f"Bearer {COMETAPI_KEY}",
    "Content-Type": "application/json",
}

# ============================================================
# Step 1: Create Video Generation Task
# ============================================================
print("Step 1: Creating video generation task...")

create_payload = {
    "model": "grok-imagine-video",
    "prompt": "A glowing crystal-powered rocket launching from the red dunes of Mars, ancient alien ruins lighting up in the background as it soars into a sky full of unfamiliar constellations",
    "duration": 10,
    "aspect_ratio": "16:9",
    "resolution": "720p",
}

create_response = requests.post(
    f"{BASE_URL}/videos/generations", headers=headers, json=create_payload
)

create_result = create_response.json()
print(f"Create response: {create_result}")

# Extract task ID from the response
task_id = create_result.get("request_id") or create_result.get("id")
if not task_id:
    print("Error: Failed to get task ID from response")
    exit(1)

print(f"Task ID: {task_id}")

# ============================================================
# Step 2: Poll for Task Status
# ============================================================
print("
Step 2: Polling task status...")

while True:
    query_response = requests.get(
        f"{BASE_URL}/videos/{task_id}", headers=headers
    )
    query_result = query_response.json()

    data = query_result.get("data", {})
    status = data.get("status", "unknown")
    progress = data.get("progress", "0%")
    print(f"Status: {status}, Progress: {progress}")

    if status in ["FAILURE", "failed"]:
        print("Video generation failed!")
        print(f"Reason: {data.get('fail_reason')}")
        exit(1)

    if status == "SUCCESS":
        video_url = data.get("data", {}).get("video", {}).get("url")
        print(f"Video generation completed!")
        print(f"Video URL: {video_url}")
        break

    time.sleep(10)

Python Code Example

import os
import time
import requests

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com/grok/v1"

headers = {
    "Authorization": f"Bearer {COMETAPI_KEY}",
    "Content-Type": "application/json",
}

# ============================================================
# Step 1: Create Video Generation Task
# ============================================================
print("Step 1: Creating video generation task...")

create_payload = {
    "model": "grok-imagine-video",
    "prompt": "A glowing crystal-powered rocket launching from the red dunes of Mars, ancient alien ruins lighting up in the background as it soars into a sky full of unfamiliar constellations",
    "duration": 10,
    "aspect_ratio": "16:9",
    "resolution": "720p",
}

create_response = requests.post(
    f"{BASE_URL}/videos/generations", headers=headers, json=create_payload
)

create_result = create_response.json()
print(f"Create response: {create_result}")

# Extract task ID from the response
task_id = create_result.get("request_id") or create_result.get("id")
if not task_id:
    print("Error: Failed to get task ID from response")
    exit(1)

print(f"Task ID: {task_id}")

# ============================================================
# Step 2: Poll for Task Status
# ============================================================
print("\nStep 2: Polling task status...")

while True:
    query_response = requests.get(
        f"{BASE_URL}/videos/{task_id}", headers=headers
    )
    query_result = query_response.json()

    data = query_result.get("data", {})
    status = data.get("status", "unknown")
    progress = data.get("progress", "0%")
    print(f"Status: {status}, Progress: {progress}")

    if status in ["FAILURE", "failed"]:
        print("Video generation failed!")
        print(f"Reason: {data.get('fail_reason')}")
        exit(1)

    if status == "SUCCESS":
        video_url = data.get("data", {}).get("video", {}).get("url")
        print(f"Video generation completed!")
        print(f"Video URL: {video_url}")
        break

    time.sleep(10)

JavaScript Code Example

// Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
const api_key = process.env.COMETAPI_KEY || "<YOUR_COMETAPI_KEY>";
const base_url = "https://api.cometapi.com/grok/v1";

const headers = {
  "Authorization": `Bearer ${api_key}`,
  "Content-Type": "application/json",
};

// ============================================================
// Step 1: Create Video Generation Task
// ============================================================
console.log("Step 1: Creating video generation task...");

const createPayload = {
  model: "grok-imagine-video",
  prompt: "A glowing crystal-powered rocket launching from the red dunes of Mars, ancient alien ruins lighting up in the background as it soars into a sky full of unfamiliar constellations",
  duration: 10,
  aspect_ratio: "16:9",
  resolution: "720p",
};

const createResponse = await fetch(`${base_url}/videos/generations`, {
  method: "POST",
  headers: headers,
  body: JSON.stringify(createPayload),
});

const createResult = await createResponse.json();
console.log("Create response:", JSON.stringify(createResult, null, 2));

const taskId = createResult?.request_id || createResult?.id;
if (!taskId) {
  console.log("Error: Failed to get task ID from response");
  process.exit(1);
}

console.log(`Task ID: ${taskId}`);

// ============================================================
// Step 2: Poll for Task Status
// ============================================================
console.log("\nStep 2: Polling task status...");

const sleep = (ms) => new Promise((resolve) => setTimeout(resolve, ms));

while (true) {
  const queryResponse = await fetch(`${base_url}/videos/${taskId}`, {
    method: "GET",
    headers: headers,
  });

  const queryResult = await queryResponse.json();
  const data = queryResult?.data || {};
  const status = data?.status || "unknown";
  const progress = data?.progress || "0%";
  console.log(`Status: ${status}, Progress: ${progress}`);

  if (["FAILURE", "failed"].includes(status)) {
    console.log("Video generation failed!");
    console.log(`Reason: ${data?.fail_reason}`);
    process.exit(1);
  }

  if (status === "SUCCESS") {
    const videoUrl = data?.data?.video?.url;
    console.log("Video generation completed!");
    console.log(`Video URL: ${videoUrl}`);
    break;
  }

  await sleep(10000);
}

Curl Code Example

#!/bin/bash

# Get your CometAPI key from https://api.cometapi.com/console/token
# Export it as: export COMETAPI_KEY="your-key-here"

BASE_URL="https://api.cometapi.com/grok/v1"

# ============================================================
# Step 1: Create Video Generation Task
# ============================================================
echo "Step 1: Creating video generation task..."

CREATE_RESPONSE=$(curl -s --location --request POST "${BASE_URL}/videos/generations" \
  --header "Authorization: Bearer $COMETAPI_KEY" \
  --header "Content-Type: application/json" \
  --data-raw '{
    "model": "grok-imagine-video",
    "prompt": "A glowing crystal-powered rocket launching from the red dunes of Mars, ancient alien ruins lighting up in the background as it soars into a sky full of unfamiliar constellations",
    "duration": 10,
    "aspect_ratio": "16:9",
    "resolution": "720p"
  }')

echo "Create response: $CREATE_RESPONSE"

# Extract task ID using jq (install with: brew install jq)
TASK_ID=$(echo "$CREATE_RESPONSE" | jq -r '.request_id // .id')

if [ "$TASK_ID" == "null" ] || [ -z "$TASK_ID" ]; then
  echo "Error: Failed to get task ID from response"
  exit 1
fi

echo "Task ID: $TASK_ID"

# ============================================================
# Step 2: Poll for Task Status
# ============================================================
echo ""
echo "Step 2: Polling task status..."

while true; do
  QUERY_RESPONSE=$(curl -s --location --request GET "${BASE_URL}/videos/${TASK_ID}" \
    --header "Authorization: Bearer $COMETAPI_KEY")

  STATUS=$(echo "$QUERY_RESPONSE" | jq -r '.data.status')
  PROGRESS=$(echo "$QUERY_RESPONSE" | jq -r '.data.progress')
  echo "Status: $STATUS, Progress: $PROGRESS"

  if [ "$STATUS" == "FAILURE" ] || [ "$STATUS" == "failed" ]; then
    echo "Video generation failed!"
    echo "$QUERY_RESPONSE" | jq -r '.data.fail_reason'
    exit 1
  fi

  if [ "$STATUS" == "SUCCESS" ]; then
    VIDEO_URL=$(echo "$QUERY_RESPONSE" | jq -r '.data.data.video.url')
    echo "Video generation completed!"
    echo "Video URL: $VIDEO_URL"
    break
  fi

  sleep 10
done