Kling 3.0 4K Image to Video API

kwaivgi/kling-v3.0-4k/image-to-video

Kling 3.0 4K Image to Video turns still images into native 4K footage, pairing frame guidance with native audio. Set the opening and closing images, then direct movement between them with scene or shot prompts.

Input

Multi Shots

Prompt(required)

497/2,500

image_urls

Drag and drop files, paste from clipboard, or provide a URL. Add images in order: the first is the first frame; the second is the optional last frame.

1/2 files added

Kling Elements (optional)

Enter an array of element objects and reference them with @element_name. Object fields are preserved in the request.

Duration

Sound

Output

Video ready

Explore another workflow

Kling 3.0 4K Text to Video

REST API

Quick Start

Submit a Kling 3.0 4K Image to Video request and use its task identifier to retrieve the video.

Authentication

Send your Vidgo API key in the Authorization header as a Bearer token.

Submit endpoint: POSThttps://api.vidgo.ai/api/generate/submit
Authentication: Authorization: Bearer YOUR_API_KEY

Submit Request

Send the model identifier and input object to the submission endpoint. Image examples contain a sample start frame; replace it with your own image URL.

Request example

REQUEST_BODY=$(cat <<'JSON'
{
  "model": "kwaivgi/kling-v3.0-4k/image-to-video",
  "input": {
    "prompt": "The camera slowly moves toward the subject. Soft daylight illuminates the scene as a gentle breeze moves through it.",
    "duration": 5,
    "multi_shots": false,
    "sound": true,
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-v3.0-4k/image-to-video/v1/01/input.png"
    ]
  }
}
JSON
)

RESPONSE=$(curl --silent --show-error --fail-with-body \
  --request POST \
  --url "https://api.vidgo.ai/api/generate/submit" \
  --header "Authorization: Bearer $VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data "$REQUEST_BODY")

CODE=$(printf '%s' "$RESPONSE" | jq -r '.code // empty')
if [ "$CODE" != "0" ] && [ "$CODE" != "200" ]; then
  printf 'API error: %s
' "$RESPONSE" >&2
  exit 1
fi
printf '%s
' "$RESPONSE"

Query Task Status

Replace {task_id} in the status URL with the identifier returned on submission.

Status endpoint

https://api.vidgo.ai/api/generate/status/{task_id}

Query this URL with the returned task_id. Continue until finished or failed; a completed video is listed in data.files.

not_startedrunningfinishedfailed

Submission response

{
  "code": 200,
  "data": {
    "task_id": "example-task-id",
    "status": "not_started",
    "created_time": "2026-09-27T00:00:00Z"
  }
}

Completed response

{
  "code": 200,
  "data": {
    "task_id": "example-task-id",
    "status": "finished",
    "created_time": "2026-09-27T00:00:00Z",
    "files": [
      {
        "file_type": "video",
        "file_url": "https://example.com/generated-video.mp4"
      }
    ]
  }
}

Complete Example

When status is finished, read file_url from video entries in files. Task IDs and output URLs below illustrate the response format.

Complete Example

set -euo pipefail
: "${VIDGO_API_KEY:?Set VIDGO_API_KEY in your environment}"

REQUEST_BODY=$(cat <<'JSON'
{
  "model": "kwaivgi/kling-v3.0-4k/image-to-video",
  "input": {
    "prompt": "The camera slowly moves toward the subject. Soft daylight illuminates the scene as a gentle breeze moves through it.",
    "duration": 5,
    "multi_shots": false,
    "sound": true,
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-v3.0-4k/image-to-video/v1/01/input.png"
    ]
  }
}
JSON
)

SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  --request POST \
  --url "https://api.vidgo.ai/api/generate/submit" \
  --header "Authorization: Bearer $VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data "$REQUEST_BODY")

TASK_ID=$(printf '%s' "$SUBMIT_RESPONSE" | jq -r '.data.task_id // .task_id // empty')
BUSINESS_CODE=$(printf '%s' "$SUBMIT_RESPONSE" | jq -r '.code // empty')
if [ "$BUSINESS_CODE" != "0" ] && [ "$BUSINESS_CODE" != "200" ]; then
  printf 'Submit failed:
%s
' "$SUBMIT_RESPONSE" >&2
  exit 1
fi
if [ -z "$TASK_ID" ]; then
  printf 'Submit response did not include task_id:
%s
' "$SUBMIT_RESPONSE" >&2
  exit 1
fi

START_TIME=$(date +%s)
POLL_DELAY=2
while true; do
  if [ $(( $(date +%s) - START_TIME )) -ge 600 ]; then
    printf 'Timed out after 600 seconds
' >&2
    exit 1
  fi
  STATUS_RESPONSE=$(curl --silent --show-error --fail-with-body \
    --url "https://api.vidgo.ai/api/generate/status/$TASK_ID" \
    --header "Authorization: Bearer $VIDGO_API_KEY")
  STATUS=$(printf '%s' "$STATUS_RESPONSE" | jq -r '.data.status // .status // empty')
  BUSINESS_CODE=$(printf '%s' "$STATUS_RESPONSE" | jq -r '.code // empty')
  if [ "$BUSINESS_CODE" != "0" ] && [ "$BUSINESS_CODE" != "200" ]; then
    printf 'Status request failed:
%s
' "$STATUS_RESPONSE" >&2
    exit 1
  fi

  case "$STATUS" in
    finished)
      printf '%s' "$STATUS_RESPONSE" | jq -r '(.data.files // .files // [])[]?.file_url'
      break
      ;;
    failed)
      printf '%s' "$STATUS_RESPONSE" | jq -r '.data.error_message // .error_message // "Generation failed"' >&2
      exit 1
      ;;
    not_started|running)
      sleep "$POLL_DELAY"
      if [ "$POLL_DELAY" -lt 10 ]; then POLL_DELAY=$((POLL_DELAY + 1)); fi
      ;;
    *)
      printf 'Unexpected task status: %s
' "$STATUS" >&2
      exit 1
      ;;
  esac
done

Request Parameters

Choose a single prompt or a sequence of shots, then set duration and sound for this endpoint.

Field	Type	Requirement	Default	Usage
image_urls	string[]	Required	Set explicitly	Provide the first frame at index 0 and an optional last frame at index 1. Use public HTTP(S) image URLs. Both single-shot and multi-shot requests accept 1–2 frames.
prompt	string	When multi_shots=false	Set explicitly	Describe movement from the reference frame in 1–2,500 characters. For multiple shots, write each scene in multi_prompt.
multi_shots	boolean	Optional	false	Choose false for one prompt, or true for separately timed shots with multi_prompt and sound=true.
multi_prompt	object[]	When multi_shots=true	Set explicitly	Add at least one object with prompt and duration. Each prompt uses 1–2,500 characters; each shot lasts 1–12 integer seconds. Shot times add up to duration.
duration	integer	Required	Set explicitly	Set the total clip length to an integer from 3 to 15 seconds. For multiple shots, use the sum of their durations.
sound	boolean	Required	Set explicitly	Use true for generated audio. A silent single shot uses false. Multiple shots use true.
kling_elements	object[]	Optional	Omit when unused	Provide reference element objects together with image_urls and refer to named elements in the prompt with @element_name. Use the documented 4K element object structure in the JSON editor.

Response Fields

Submission returns a task identifier. Query its status to retrieve generated video URLs.

Field	Type	Usage
code	integer	Response code: 0 or 200 indicates a successful API operation.
data.task_id	string	Identifier returned on submission; use it to query the same task.
data.status	string	One of not_started, running, finished or failed. A successful submission is followed by status checks.
data.created_time	string	Task creation time returned by the service.
data.files	array	Generated media entries returned when the task finishes.
data.files[].file_type	string	Use entries with file_type=video for the generated clip.
data.files[].file_url	string	Video URL for playback or download.
data.error_message	string \| null	Read the task error detail when status=failed.

Task Lifecycle

Keep the returned task_id and follow status until finished or failed.

not_started
The request is accepted and waiting to start. Keep its task_id for the next status check.
running
The video is being generated. Continue checking the same task.
finished
Generation is complete. Read the video entries in files.
failed
Generation ended with an error. Review error_message before submitting an adjusted request.

Error Handling

Check timing before submittingFor multiple shots, use sound=true and ensure individual durations sum to a total of 3–15 seconds.
Keep reference URLs accessibleUse public HTTP(S) media URLs that remain reachable while the task is processed.
Resume interrupted status checksIf a status request fails, retry the status check using the same task_id. Submit a new generation only when you intend to create another clip.
Read the returned errorFor a failed task, inspect error_message and adjust the relevant input before trying again.

Specifications

Specification	Value	Usage
Input mode	Image to Video	A start frame plus motion direction; an end frame is optional.
Resolution	4K	Native 4K video output.
Clip duration	3–15 seconds	Whole seconds; each shot in a sequence uses 1–12 seconds.
Audio	sound	Generate audio with true; single-shot silent generation uses false.
Model identifier	kwaivgi/kling-v3.0-4k/image-to-video	Public model value for this request.

View the Kling 3.0 4K Image to Video API docs

Kling 3.0 4K Image to Video

Kling 3.0 4K Image to Video is Kuaishou’s model for animating an existing visual into native 4K footage. Start and end frames establish visual destinations, while shot prompts set the action and joint audio generation adds the soundtrack. It suits detailed product imagery, architectural scenes and campaign visuals that will be assembled or presented in a 4K workflow.

Why Choose Kling 3.0 4K Image to Video

Create for a 4K timelineGenerate native 4K footage for detailed scenes and large-screen presentation concepts.
Start from the chosen visualUse the first frame to establish the subject and composition, then describe the movement that brings it to life.
Give the ending a destinationPair a start frame with an end frame to guide a single shot toward a planned closing image.
Anchor recurring subjectsName reference elements and call them out in the prompt to guide the appearance of a character or product.

Parameters

Parameter	When to use	Usage
image_urls	Required	Provide the first frame at index 0 and an optional last frame at index 1. Use public HTTP(S) image URLs. Both single-shot and multi-shot requests accept 1–2 frames. Default`Set explicitly`
prompt	When multi_shots=false	Describe movement from the reference frame in 1–2,500 characters. For multiple shots, write each scene in multi_prompt. Default`Set explicitly`
multi_shots	Optional	Choose false for one prompt, or true for separately timed shots with multi_prompt and sound=true. Default`false`
multi_prompt	When multi_shots=true	Add at least one object with prompt and duration. Each prompt uses 1–2,500 characters; each shot lasts 1–12 integer seconds. Shot times add up to duration. Default`Set explicitly`
duration	Required	Set the total clip length to an integer from 3 to 15 seconds. For multiple shots, use the sum of their durations. Default`Set explicitly`
sound	Required	Use true for generated audio. A silent single shot uses false. Multiple shots use true. Default`Set explicitly`
kling_elements	Optional	Provide reference element objects together with image_urls and refer to named elements in the prompt with @element_name. Use the documented 4K element object structure in the JSON editor. Default`Omit when unused`

How to Use

Prepare the starting sceneAdd a start frame and describe the movement. Add an end frame for a planned finish.
Choose the scene structureUse one prompt for a single shot. For multiple shots, enable Multi Shots, write each shot and set its duration. Use a start frame and an optional end frame in both single-shot and multi-shot requests.
Review timing and soundSet a total of 3–15 seconds. Multiple shots keep Sound enabled; review the displayed estimate before running.
Run and review the clipSelect Run, then preview the completed video. Refine the action or camera direction for the next version.

Pricing

Kling 3.0 4K Image to Video is billed by video duration. Multiply seconds by the rate for your Sound setting.

Generation mode	Per second	Billing basis
Silent single shot	$0.250/second	sound=false and multi_shots=false. A 5-second clip costs $1.250.
Single or multiple shots with audio	$0.250/second	sound=true. A 5-second clip costs $1.250. Multiple shots use their combined duration.

Best Use Cases

4K campaign visualsBring a detailed campaign still into a 4K editing workflow, using opening and closing frames to plan the visual arc.
Character momentsStart with a character image and describe a glance, turn or gesture, using element references for visual identity.
Animated key visualsDevelop an existing campaign image into a short scene with controlled movement and sound.

Pro Tips

Choose a start frame with a clearly visible subject and the composition you want to animate.
Describe what moves and how the camera moves; let the reference image carry the static appearance.
For a single shot, choose an end frame that follows naturally from the opening action.
Give each reference element a distinct name and use that exact @name in the prompt.

Usage Notes

Use a start frame and an optional end frame in both single-shot and multi-shot requests.
Keep the total duration between 3 and 15 seconds and enable Sound for multiple shots.

Related Models

Gemini Omni 1.1 Flash Text-to-Video preview

Gemini Omni 1.1 Flash Text-to-VideoGoogle Gemini Omni 1.1 Flash video model with native audio, text-to-video, first/last-frame image-to-video, reference-image video, and 3–10 second video edit at 360p to 4K.

Gemini Omni 1.1 Flash Image-to-Video preview

Gemini Omni 1.1 Flash Image-to-VideoGoogle Gemini Omni 1.1 Flash video model with native audio, text-to-video, first/last-frame image-to-video, reference-image video, and 3–10 second video edit at 360p to 4K.

Gemini Omni 1.1 Flash Reference-to-Video preview

Gemini Omni 1.1 Flash Reference-to-VideoGoogle Gemini Omni 1.1 Flash video model with native audio, text-to-video, first/last-frame image-to-video, reference-image video, and 3–10 second video edit at 360p to 4K.

Gemini Omni 1.1 Flash Video Edit preview

Gemini Omni 1.1 Flash Video EditGoogle Gemini Omni 1.1 Flash video model with native audio, text-to-video, first/last-frame image-to-video, reference-image video, and 3–10 second video edit at 360p to 4K.

MiniMax H3 Max Turbo Text to Video preview

MiniMax H3 Max Turbo Text to VideoMiniMax H3 Max Turbo Text to Video turns text into coherent 5–15 second videos with synchronized native stereo audio and fast generation. It follows physical motion while rendering subtle changes in light and shadow.

MiniMax H3 Max Turbo Image to Video preview

MiniMax H3 Max Turbo Image to VideoMiniMax H3 Max Turbo Text to Video turns text into coherent 5–15 second videos with synchronized native stereo audio and fast generation. It follows physical motion while rendering subtle changes in light and shadow.

MiniMax H3 Max Text to VideoMiniMax H3 Max Text to Video turns text prompts into cinematic video with native stereo audio, faster-than-real-time generation, and precise instruction following. It renders complex camera work and physical motion with fine detail across multiple resolutions and common aspect ratios.

MiniMax H3 Max Image to VideoMiniMax H3 Max Text to Video turns text prompts into cinematic video with native stereo audio, faster-than-real-time generation, and precise instruction following. It renders complex camera work and physical motion with fine detail across multiple resolutions and common aspect ratios.

MiniMax H3 Max Reference to Video preview

MiniMax H3 Max Reference to VideoMiniMax H3 Max Text to Video turns text prompts into cinematic video with native stereo audio, faster-than-real-time generation, and precise instruction following. It renders complex camera work and physical motion with fine detail across multiple resolutions and common aspect ratios.

Wan 3.0 Prime Text-to-VideoAlibaba Wan 3.0 Prime video model with text-to-video, image-to-video, and reference-to-video workflows at 480p, 720p, or 1080p for 2 to 30 seconds with optional synchronized audio.

Wan 3.0 Prime Image-to-VideoAlibaba Wan 3.0 Prime video model with text-to-video, image-to-video, and reference-to-video workflows at 480p, 720p, or 1080p for 2 to 30 seconds with optional synchronized audio.

Wan 3.0 Prime Reference-to-Video preview

Wan 3.0 Prime Reference-to-VideoAlibaba Wan 3.0 Prime video model with text-to-video, image-to-video, and reference-to-video workflows at 480p, 720p, or 1080p for 2 to 30 seconds with optional synchronized audio.

Wan 3.0 Text-to-VideoAlibaba Tongyi Lab's Wan 3.0 multimodal video model with separate text-to-video, first/last-frame image-to-video, and reference-to-video workflows at 480p, 720p, or 1080p for 2 to 30 seconds.

Wan 3.0 Image-to-VideoAlibaba Tongyi Lab's Wan 3.0 multimodal video model with separate text-to-video, first/last-frame image-to-video, and reference-to-video workflows at 480p, 720p, or 1080p for 2 to 30 seconds.

Wan 3.0 Reference-to-VideoAlibaba Tongyi Lab's Wan 3.0 multimodal video model with separate text-to-video, first/last-frame image-to-video, and reference-to-video workflows at 480p, 720p, or 1080p for 2 to 30 seconds.

FLUX 3 Text to VideoBlack Forest Labs FLUX 3 video model with text-to-video, image-to-video, first-last-frame, extend-video, and keyframes workflows at 720p or 1080p for 5 to 20 seconds, with optional native audio.

Kling 3.0 4K Image to Video API frequently asked questions

What is the Kling 3.0 4K Image to Video API?

Kling 3.0 4K Image to Video is a Kuaishou model for animating still images. It creates native 4K footage with audio and frame guidance. Image conditioning guides the starting composition, while element references give recurring subjects a visual anchor. You can call it programmatically or try it from the playground above.

Can Kling 3.0 4K Image to Video use an end frame with multiple shots?

Yes. This 4K endpoint accepts a start frame at image_urls[0] and an optional end frame at image_urls[1] in multi-shot requests. Enable sound and multi_shots, then use per-shot prompts and timings to describe the route between the opening and closing images.

Does Kling 3.0 4K Image to Video animate images at native 4K?

Yes. The output tier is native 4K. Start with an image that clearly shows the subject and its important details, and describe the desired movement so the animation follows the visual direction of the source.

How do timings guide Kling 3.0 4K Image to Video transitions?

Give each shot a prompt and 1–12 integer seconds. Their combined duration must be 3–15 seconds, with multi_shots=true and sound=true. Allocate time to the action that leads toward the closing frame, such as a camera pullback or a subject turn.

How do I add audio in Kling 3.0 4K Image to Video?

Set sound=true and describe the desired dialogue, ambient sound or effects alongside the visual action. Single shots can use sound=false for silent footage; multi-shot sequences use sound=true.

How do element references work in Kling 3.0 4K Image to Video?

Supply kling_elements alongside image_urls using the 4K element structure in the API documentation. Fill the element objects in JSON and use @element_name in prompts to point to named subjects. Keep the start frame as the opening visual reference.

When should I choose Kling 3.0 4K over Kling 3.0 Pro?

Choose 4K for native 4K footage intended for detailed scenes or a 4K timeline. Pro delivers 1080p at a lower per-second rate; select the tier that matches your final output.

How does Sound affect Kling 3.0 4K billing?

Both sound=true and sound=false cost $0.250 per second. Multiply the rate by total clip duration; a five-second clip costs $1.250. Multiple shots keep sound=true.