kling / v2-1-master-text-to-video
Commercial usePart of the Kling API family — compare all Kling models →

The Kling 2.1 model powers cutting-edge video generation with hyper-realistic motion, advanced physics, and high-resolution outputs up to 1080p. Its enhanced semantic understanding and fast rendering make it ideal for dynamic, professional-grade video creation.

Model Type:
Pricing: A 5-second video costs 160 credits ($0.80), and a 10-second video costs 320 credits ($1.60). High-tier top-ups (+10% bonus) bring effective pricing down to ~$0.72 for a 5-second video and ~$1.44 for a 10-second video.
Input

The text prompt describing the video you want to generate

0/5000

The duration of the generated video in seconds

The aspect ratio of the generated video frame

Elements to avoid in the generated video

The CFG (Classifier Free Guidance) scale is a measure of how close you want the model to stick to your prompt

A configurable parameter. Defaults to true in the Playground.

There is no guarantee that everything can be filtered out; if you are not satisfied with the results, you will need to make your own arrangements.
Output
output typevideo
Examples

Explore different use cases and parameter configurations

README

Complete guide to using Kling V2.1

Kling 2.1 API: Standard, Pro and Master Video Generation

Kling 2.1 is Kuaishou's video model series, released at the end of May 2025 with a Standard mode, a High Quality mode and a Master Edition. On Kie.ai it runs as four API models: Standard, Pro and Master image-to-video, and Master text-to-video.

Get Kling API Key
Hero section demo image showing interface components

Kling 2.1 API Parameters on Kie.ai

All four models use POST /api/v1/jobs/createTask, and results come back by callBackUrl or Get Task Details. There is no resolution parameter: the output size depends on the model.

ParameterStandardProMaster
Model on Kie.aikling/v2-1-standardkling/v2-1-prokling/v2-1-master-image-to-video, kling/v2-1-master-text-to-video
Text-to-videoNoNoYes (kling/v2-1-master-text-to-video)
Start image (image_url)RequiredRequiredRequired for image-to-video
End image (tail_image_url)NoOptionalNo
Duration"5" or "10" seconds"5" or "10" seconds"5" or "10" seconds
Aspect ratioFollows the start imageFollows the start imageImage-to-video: follows the start image. Text-to-video: 16:9, 9:16 or 1:1
Output size in our tests on Kie.ai960×960 from a square image1440×1440 from a square image1440×1440 from a square image; 1920×1080 at 16:9 and 1080×1920 at 9:16 (text-to-video)
Audio track in our tests on Kie.aiNoneNoneNone
Prompt / negative promptUp to 5,000 / 500 charactersUp to 5,000 / 500 charactersUp to 5,000 / 500 characters
cfg_scale0 to 1, step 0.1, default 0.50 to 1, step 0.1, default 0.50 to 1, step 0.1, default 0.5
Input imagesHosted JPEG or PNG URL, up to 10 MBHosted JPEG or PNG URLs, up to 10 MB eachHosted JPEG or PNG URL, up to 10 MB
PricingShown in the pricing card aboveShown in the pricing card aboveShown in the pricing card above

Kling 2.1 Standard vs Pro vs Master: Which One to Use

Kuaishou describes the Kling 2.1 series as having a Standard (720p) mode and a High Quality (1080p) mode, plus a premium Master Edition that it says delivers superior motion performance and enhanced semantic responsiveness. To compare the three on Kie.ai, we ran Standard, Pro and Master image-to-video twice in October 2026 with the same square image and a 10-second duration. With a simple prompt (the robot waves while the camera orbits), all three looked alike. With a hard one (the robot breakdances while the camera circles quickly), each did something different at the 5-second mark shown in the image: Standard held a handstand, Pro was mid-kick and Master was mid-spin with visible motion blur, which it also showed in its other high-motion frames. From still frames we could not tell whether one model handles fast motion better, and this is one run per model, so test your own prompts before you decide. On LMArena's image-to-video leaderboard (September 2026), Master and Standard are within each other's range. A practical default follows the model list: Standard or Pro for image-to-video, Pro when you need an end frame, and Master when you need text-to-video. Pricing for each model is shown in the pricing card above.

Kling 2.1 Standard vs Pro vs Master: Which One to Use
StandardProMaster
Text-to-videoNoNoYes
End frameNoYesNo
Output size from the same square image (our test)Smaller (see the parameter table)LargerLarger
Simple motion: waving, slow orbit (our test)Followed the promptFollowed the promptFollowed the prompt
Breakdance prompt, frame at 5 s (our test)HandstandMid-kickMid-spin, with motion blur
LMArena image-to-video score (Sep 21, 2026)1229±8Not listed1235±8

What Kling 2.1 Can Generate

Each image below shows three frames (start, middle and end) from one video we generated with Kling 2.1 on Kie.ai in October 2026 from the prompt described in its caption; frames are shown as generated, resized for the web.

Text-to-Video in 16:9: Waves on Volcanic Rocks at Sunset

Kling 2.1 Master turned a one-sentence prompt into a 5-second landscape video: waves crash against black volcanic rocks, the spray catches the low sun, and the camera moves forward over the rocks as the prompt asked.

Text-to-Video in 16:9: Waves on Volcanic Rocks at Sunset

Text-to-Video in 9:16: Pouring Latte Art

A 10-second vertical video from Kling 2.1 Master: milk is poured into a cup and a layered pattern forms on the surface while the camera pushes in. The prompt asked for a leaf pattern and the result is closer to a tulip, so check details like this against your brief.

Text-to-Video in 9:16: Pouring Latte Art

Start and End Frames on Pro: From a Studio to a Library

Kling 2.1 Pro takes an optional end image. Given the robot on a plain background as the start frame and the same robot reading in a library armchair as the end frame, it generated a 5-second walk from one scene into the other, and the last frame matches the end image.

Start and End Frames on Pro: From a Studio to a Library
4.9/ 5
Top rated · 23,862 ratings
Add your rating

How to Use the Kling 2.1 API on Kie.ai

  • Step 1. Get Your API Key

    Sign up on Kie.ai, open the API Key Management page and create a key. Every request sends it in the Authorization header.

  • Step 2. Test Your Prompt in the Playground

    You can try this model in the playground on this page without writing code. Start with a short prompt and a 5-second duration, then switch models to compare.

  • Step 3. Create a Task and Get the Result

    Send a POST request to /api/v1/jobs/createTask with the model name and an input object holding your prompt, duration and, for image-to-video, image_url, using the parameters in the table above. Pass a callBackUrl to receive the result when the task finishes, verify each callback with the HMAC signature described in the Kie.ai webhook guide, and query Get Task Details for the authoritative result.

Tips Before You Ship

Send Duration as a String

duration is required on every model and takes the string "5" or "10", not a number.

Choose the Model Before You Build the Request

Standard, Pro and Master image-to-video all need image_url, and their videos follow the shape of that image. Only kling/v2-1-master-text-to-video works from text alone, and it is the only model that takes aspect_ratio.

Plan for Silent Videos

None of our test videos on Kie.ai came with sound, so add music or sound effects in a separate step.

Pass Image URLs, Not Files

image_url and tail_image_url take hosted image URLs, not file content. Files that are not hosted yet can be uploaded first with the Kie.ai File Upload API (Base64, file stream or remote URL).

Why Use Kling 2.1 on Kie.ai

Same Key Across the Kie.ai Market

All models in the Kie.ai market use the same authentication method: an Authorization: Bearer YOUR_API_KEY header with the key from your Kie.ai API Key Management page.

Free Credits to Start

New Kie.ai users receive 80 free credits for testing. Your account has one credit balance that all generation services draw from; check it with the Get Remaining Credits API.

Nothing to Host or Install

Kie.ai provides this model as an API with a playground; there is no app to download or install. There are no GPUs or model weights to manage.

FAQs About Kling 2.1 API

Pricing for each model and duration is shown in the pricing card above.

Not on Kie.ai: each video uses credits, as shown in the pricing card above. New Kie.ai users receive 80 free credits for testing.

Yes. Kuaishou runs its own Kling AI developer platform, and on Kie.ai Kling 2.1 is available as four API models: kling/v2-1-standard, kling/v2-1-pro, kling/v2-1-master-image-to-video and kling/v2-1-master-text-to-video.

On Kie.ai, only Master does text-to-video and only Pro accepts an end frame. From the same square image, Pro and Master returned larger videos than Standard in our test; the comparison table above shows the three side by side.

cfg_scale sets how closely the video sticks to your prompt. On Kie.ai it takes a value from 0 to 1 in steps of 0.1, and the default is 0.5.

There is no resolution setting: the output size depends on the model and, for image-to-video, on the shape of your image. Kuaishou describes a Standard (720p) mode and a High Quality (1080p) mode; the parameter table above lists the sizes we measured on Kie.ai.

Not in our tests on Kie.ai: in October 2026 each of the four models returned a video track only, with no audio track.

Yes, on Pro: pass the start image in image_url and the end image in tail_image_url. Standard and Master image-to-video take a start image only. The studio-to-library example above was made this way.

Each request makes a 5-second or 10-second video on every model; set duration to "5" or "10".

Create one on the API Key Management page after you sign up on Kie.ai. The steps above show where the key goes in a request.