Start building

Kling Video V3

kwaivgi/kling-video/v3

Kling Video V3 is KwaiVGI's unified video generation model. Generate or transform videos from prompts, images, reference videos, and multi-shot text while choosing standard or pro mode per request.

Use in agentGet API key
Base price
$0.63USD / run
Execution
async
Model type
video
Input fields
13

Try the model

Playground

Open playground
Input
Text prompt for video generation or transformation. When using reference media, refer to them as @Image1, @Image2, and @Video1.
Generation quality tier. Allowed values: standard, pro, 4k.

PNG, JPEG, WebP, or GIF · 20 MiB maximum

Optional first-frame image URL for image-to-video.
Optional reference video URL for video-to-video generation.
Allowed values: image, video.

PNG, JPEG, WebP, or GIF · 20 MiB maximum

Optional end-frame image URL for image-to-video.

PNG, JPEG, WebP, or GIF · 20 MiB maximum each

Optional reference image URLs.
Generated video duration in seconds. Allowed values: 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15.
Allowed values: 16:9, 9:16, 1:1.
Allowed values: customize, intelligent.
Enter a JSON array.
OutputReady

Your output will appear here

Complete the inputs, then click Run.

Specifications

Pricing

Base price
$0.63 / run
Billing formula
params.resolution == "4k" ? params.duration * 0.42 : (params.resolution == "pro" ? params.duration * 0.168 : params.duration * 0.126)

Context & modalities

Input
Schema-defined
Output
video

Capabilities

Chat
Not supported
Vision
Not supported
Reasoning
Not supported
Structured output
Not supported
Function calling
Not supported
Audio input
Not supported

Access

Provider
KwaiVGI
Model ID
kwaivgi/kling-video/v3
Execution
async
API
Unified Run API
Endpoint
/v1/run

API README

Kling Video V3

Kling Video V3 is a V3 endpoint in the Kling family. It is built for creators who need to turn a concrete creative brief into a controlled visual sequence: the request establishes the source material, the intended subject behavior, the camera language, and the atmosphere of the finished shot. The standard route keeps that workflow explicit instead of hiding its input assumptions behind a generic video-generation label.

In practice, this route accepts image, video, end image as creative context and exposes duration, resolution, aspect ratio, generate audio for delivery planning. That makes it suitable for shot-based pipelines where teams must preserve a source, direct a transformation, or control the final format without losing sight of the model's central task. Write prompts as a compact shot plan—subject, action, setting, camera, light, and timing—then use the structured fields for constraints that should remain deterministic across iterations.

Highlights

  • Prompt-to-scene generation. Builds the shot from written direction, including subject action, environment, lighting, lens language, and pacing.
  • Synchronized audio path. Audio conditioning or generation is available in the same request, helping sound and picture share one creative plan.
  • Endpoint composition. An optional end frame can anchor where the shot lands, giving transitions a deliberate visual destination.
  • Resolution-aware delivery. Explicit output resolution lets the request target preview or finishing needs before generation.

Pricing

The request price is calculated with params.resolution == "4k" ? params.duration * 0.42 : (params.resolution == "pro" ? params.duration * 0.168 : params.duration * 0.126).

ConfigurationPrice
duration=3, resolution=standard$0.378000
duration=3, resolution=pro$0.504000
duration=3, resolution=4k$1.260000
duration=4, resolution=standard$0.504000
duration=4, resolution=pro$0.672000
duration=4, resolution=4k$1.680000
duration=5, resolution=standard$0.630000
duration=5, resolution=pro$0.840000
duration=5, resolution=4k$2.100000
duration=6, resolution=standard$0.756000
duration=6, resolution=pro$1.008000
duration=6, resolution=4k$2.520000
duration=7, resolution=standard$0.882000
duration=7, resolution=pro$1.176000
duration=7, resolution=4k$2.940000
duration=8, resolution=standard$1.008000
duration=8, resolution=pro$1.344000
duration=8, resolution=4k$3.360000
duration=9, resolution=standard$1.134000
duration=9, resolution=pro$1.512000
duration=9, resolution=4k$3.780000
duration=10, resolution=standard$1.260000
duration=10, resolution=pro$1.680000
duration=10, resolution=4k$4.200000
duration=11, resolution=standard$1.386000
duration=11, resolution=pro$1.848000
duration=11, resolution=4k$4.620000
duration=12, resolution=standard$1.512000
duration=12, resolution=pro$2.016000
duration=12, resolution=4k$5.040000
duration=13, resolution=standard$1.638000
duration=13, resolution=pro$2.184000
duration=13, resolution=4k$5.460000
duration=14, resolution=standard$1.764000
duration=14, resolution=pro$2.352000
duration=14, resolution=4k$5.880000
duration=15, resolution=standard$1.890000
duration=15, resolution=pro$2.520000
duration=15, resolution=4k$6.300000

The model card records a base price of $0.630000; the formula above determines usage-priced requests.

When to Use

ScenarioRecommendation
Choose this routeUse it when the deliverable specifically calls for V3, rather than a neighboring generation mode.
Prepare the sourceProvide image, video in the format described by the request schema.
Direct the shotDescribe the subject, action, environment, camera movement, lighting, and temporal progression in that order.
Control continuityUse endpoint frames, reference media, strength, or audio controls when those fields are available instead of burying hard constraints in prose.
Plan deliverySet duration, frame count, resolution, and aspect ratio explicitly when the schema exposes them, then compare iterations with a stable seed where supported.

Prompt Guide

For V3, describe one coherent shot rather than a list of visual keywords. Put the main subject and action first, follow with location and staging, then add camera movement, lens or framing, lighting, mood, and any timed change. Keep URLs and hard delivery choices in their dedicated fields.

{
  "prompt": "Based on @Video1, make the character from @Image1 dance.",
  "duration": 3,
  "resolution": "standard",
  "aspect_ratio": "16:9",
  "generate_audio": true,
  "image": "https://static.sandbase.ai/examples/mirrored/f77b827ca02a-TNErq9yD7ZxGRATjfAqnh_EIgJSN67.png"
}

Technical Specs

SpecificationValue
Model IDkwaivgi/kling-video/v3
Required inputsprompt
ExecutionAsynchronous generation job
Request controls13 documented fields
Outputurl, content_type

Request fields

FieldType and constraints
imagestring; Optional
videostring; Optional
imagesarray; Optional
promptstring; Required
durationinteger; Optional; Options: 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15; Default: 5
end_imagestring; Optional
shot_typestring; Optional; Options: customize, intelligent; Default: "customize"
resolutionstring; Optional; Options: standard, pro, 4k; Default: "standard"
aspect_ratiostring; Optional; Options: 16:9, 9:16, 1:1; Default: "16:9"
multi_promptarray; Optional
generate_audioboolean; Optional; Default: true
keep_original_soundboolean; Optional; Default: true
character_orientationstring; Optional; Options: image, video

Related Models

Start building

Send your first request

OpenAI-compatible endpoint with unified authentication and usage tracking.

Production API
Unified Run API endpoint
https://api.sandbase.ai/v1/run
Model ID
kwaivgi/kling-video/v3
# The quoted heredoc keeps Unicode and shell metacharacters unchanged.
result=$(curl --fail-with-body --silent \
  -X POST "https://api.sandbase.ai/v1/run" \
  -H "Authorization: Bearer $SANDBASE_API_KEY" \
  -H "Content-Type: application/json" \
  --data-binary @- <<'SANDBASE_JSON'
{
  "model": "kwaivgi/kling-video/v3",
  "image": "https://static.sandbase.ai/examples/mirrored/f77b827ca02a-TNErq9yD7ZxGRATjfAqnh_EIgJSN67.png",
  "video": "https://static.sandbase.ai/examples/mirrored/f1ecbb471a16-hklvF__w53diz6Rve7f5__JuDW2xl0mr6sJ_Kjz3Vxe_vidoeook--1-_1.mp4",
  "prompt": "Based on @Video1, make the character from @Image1 dance.",
  "duration": 5,
  "shot_type": "customize",
  "resolution": "standard",
  "aspect_ratio": "16:9",
  "generate_audio": true,
  "keep_original_sound": true,
  "character_orientation": "image"
}
SANDBASE_JSON
)
run_id=$(printf '%s' "$result" | jq -r .id)
for attempt in $(seq 1 120); do
  status=$(printf '%s' "$result" | jq -r .status)
  case "$status" in completed|failed|timeout) break ;; esac
  sleep 2
  result=$(curl --fail-with-body --silent \
    -H "Authorization: Bearer $SANDBASE_API_KEY" \
    "https://api.sandbase.ai/v1/run/$run_id")
done
status=$(printf '%s' "$result" | jq -r .status)
[ "$status" = completed ] || { echo "Generation ended: $status" >&2; exit 1; }
printf '%s\n' "$result"

Choose your model

Compare models

All language models
ModelContextInput / 1MOutput / 1MReleased
Kling Video V3This model
KwaiVGI		———Jul 6, 2026
Kling Video O3
KwaiVGI		———Jul 6, 2026
Kling 3.0 Omni Pro Video Reference
KwaiVGI		———Jul 3, 2026
Kling 3.0 Omni Pro Video Edit
KwaiVGI		———Jul 3, 2026
Kling 3.0 Omni Pro Reference to Video
KwaiVGI		———Jul 3, 2026
Kling 3.0 Omni Pro Image to Video
KwaiVGI		———Jul 3, 2026