Skip to main content

Kling 2.1 image-to-video animates an opening frame into a silent 5 or 10 second clip. Choose 720p standard or 1080p pro, and add an optional closing frame when you render at 1080p; the aspect ratio follows your image.

From $0.056/s

Workbench

Input

Image to video
209/5000
Describe only what happens after the opening frame: the action and the camera move. at least 1 characters · at most 5000 characters
1/1

Drop files or media links, paste, or browse · 1 file · Max 10 MB · JPG · PNG

  • 1
1 / 1 files
Opening frame. JPEG or PNG only.
0/1

Drop files or media links, paste, or browse · 1 file · Max 10 MB · JPG · PNG

0 / 1 files
Optional closing frame, accepted only at 1080p. A 720p request that includes it is turned down, with no charge.
affects price
Use Kling 2.1 to test a still image as a moving shot in 720p, or select 1080p when you also need the optional final image; neither setting generates a soundtrack.
Durationaffects price
510
Whole seconds, 5 or 10. Drives the price.
0/500
For Kling 2.1, list unwanted visual artifacts in negative_prompt and keep the intended action in prompt; the two fields serve different parts of the shot brief. at most 500 characters
Prompt adherence
Use guidance_scale to tune how literally Kling 2.1 interprets motion instructions: a higher value favours the wording, while a lower value leaves more freedom in the take. no lower than 0 · no higher than 1
Safety checkerOff

SpicyAPI does not run a safety checker on this model. Your prompt and media go to the model as written, and the model itself decides what comes back.

Checking your account before enabling submission and pricing.

Output

Ready for a new run
Example result

Missing the model you need?

IMAGE TO VIDEO, ENTRY PRICE

Kling 2.1 API at a glance

Use Kling 2.1 to test a still image as a moving shot in 720p, or select 1080p when you also need the optional final image; neither setting generates a soundtrack.

Resolution tiers
2
720p standard, or 1080p pro for a per-second premium.
Longest clip
10S
Kling 2.1 supports 5 or 10 second takes; choose the length for one planned action before submitting the render.
Closing frame
LAST
Available on 1080p only, to pin the ending.
Guidance scale
0-1
Use guidance_scale to tune how literally Kling 2.1 interprets motion instructions: a higher value favours the wording, while a lower value leaves more freedom in the take.
Audio track
NO
Kling 2.1 produces a silent video for your edit, so plan voiceover, music and effects as a separate post-production step.
Mode offered
1
Image-to-video only - there's no text-to-video on this generation.

Results

What you can make with Kling 2.1

  • Entry price

    An opening image, two resolution choices

    Use Kling 2.1 to test a still image as a moving shot in 720p, or select 1080p when you also need the optional final image; neither setting generates a soundtrack.

    Image-to-video, 720p standard tier: a single continuous arc from swing to guard stance.

  • 1080p pro

    A closing frame to pin the ending

    Send resolution as 1080p and add last_image_url alongside your opening frame, and the clip is built to land on that exact pose - useful for chaining several clips into one longer scene.

    Image-to-video, 1080p, opening and closing frames: the walk builds toward a specific final pose.

  • Known limit

    A dragging train doesn't stay physically consistent

    In one of our tests, a long velvet gown train visibly shortened over the course of the clip - a reminder that trailing fabric, long hair or loose props are the kind of detail this generation can lose track of across a run.

    Image-to-video, 1080p: the gown's train visibly shortens across the clip, a known limit on trailing fabric.

Content test

Is Kling 2.1 uncensored? What we tested

Content levels tested on Kling 2.1, by mode
LevelImage to video
Mild
Not tested yet
Moderate
RefusedTested Sep 28, 2026
Strong
Not tested yet
Very strong
Not tested yet
Maximum
Not tested yet
How we test

How each mode behaves

Image to video
This generation is older than Kling's 2.5 Turbo, 2.6 and 3.0 tiers: motion and small physical details (a dragging train, flowing hair) are more likely to drift or shorten across a clip, and there is no sound. A closing frame is only available at 1080p, and even then it pins the ending, not the path there.
All modes
For Kling 2.1 image animation, supplying your own frames does not remove Kling’s review of the request and result; adult NSFW requests are rejected without a generation charge. SpicyAPI adds no separate filter to Kling 2.1, and cannot disable Kling’s checks. MiniMax H3 Spicy is a separate unrestricted choice from Kling 2.1, with its own input options.
All modes
This generation offers image-to-video only - there is no text-to-video mode on 2.1. If you need to start from a written scene, use Kling 2.1 Master (the same generation's text-capable tier) or a newer Kling model.

Use cases

What to build with the Kling 2.1 API

  • An opening image, two resolution choices

    Before committing to a newer, pricier Kling tier, use 2.1's 720p standard to see whether a camera move or an action reads well from your opening frame.

  • Simple push-ins, turns and walks

    Straightforward single-subject motion - a turn toward camera, a slow dolly, a walk to a mark - is where this generation is most reliable.

  • A two-frame sequence with a locked ending

    Use 1080p with both image_url and last_image_url when the final pose matters, such as the last shot before a cut to a new scene.

  • Silent b-roll for a project scoring its own sound

    This model has no audio track at all, which is a feature rather than a limitation when you're going to add your own voiceover, music or sound design in post.

Prompts

Kling 2.1 prompt examples you can copy

  • Image to video

    Name the destination pose, not just the action

    Result from this prompt

    She turns and walks slowly back along the pier toward the camera through the drifting fog, the wind catching her red scarf, and stops close to the lens, looking straight into it. The fog thins and the dawn light warms the water behind her.

    Why it works. Pairing this prompt with a closing frame at 1080p gave the clip a specific final pose to build toward, instead of an open-ended walk.

    Try this prompt
  • Image to video

    One clean arc, one ending pose

    Result from this prompt

    She spins and whips the katana up through the rain in one clean arc, water spraying off the blade, then settles into a low guard facing the camera. Rain pours across the rooftop, the city lights glowing below, slow push-in.

    Why it works. A single continuous action described start to finish - the spin, the arc, the settle - reads more reliably on this generation than several distinct beats in one prompt.

    Try this prompt
  • Image to video

    Flag anything that has to hold its shape

    Result from this prompt

    He takes his hand off the glass, turns away from the window and walks toward the camera tying the belt of his robe, stopping with a small smile as the morning light floods the room behind him. Slow dolly back.

    Why it works. A short, single-beat action with a named camera move (slow dolly back) is the safest bet on this generation; save multi-beat, prop-heavy scenes for a newer tier.

    Try this prompt
  • Image to video

    Expect drift on trailing fabric across 10 seconds

    Result from this prompt

    She turns and climbs two more steps, the long red velvet train dragging heavily behind her across the marble stairs, then looks back over her shoulder toward the camera. The chandelier glitters above as the camera cranes slowly up.

    Why it works. In our test, the train's length visibly shortened over the clip - a known limit of this generation with long trailing fabric, not a prompt-wording problem.

    Try this prompt

Compare

Kling 2.1 vs similar models: which one to pick

ModelModesFromContentPick it when
Kling 2.1This pageImage to video$0.056 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.Use Kling 2.1 to test a still image as a moving shot in 720p, or select 1080p when you also need the optional final image; neither setting generates a soundtrack.
Kling 2.1 MasterImage to video · Text to video$0.28 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.You need the dedicated high-quality Master version of 2.1, rather than the 720p and 1080p image-animation options on this page.
Kling 2.5 TurboImage to video · Text to video$0.07 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.You want to start a 1080p shot from text as well as an image, with generated sound included and 5 or 10 seconds to stage the action.
Kling 3.0Character animation · Image to video · Text to video$0.067 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.You want Kling's flagship: 4K, an optional closing frame, switchable sound and Motion Control.

Pricing

Kling 2.1 API pricing

A 5-second clip starts at $0.281 at 720p and $0.489 at 1080p.

FAQ

Kling 2.1 API FAQ

About Kling 2.1

Kling 2.1 image-to-video starts with your image, offers 720p or 1080p, and accepts an optional closing frame at 1080p for a silent 5 or 10 second clip. Kling 2.1 runs as an asynchronous task billed in US dollars per second of output; check the selected settings before submission, and retrieve the result by task ID or webhook. For Kling 2.1 image animation, supplying your own frames does not remove Kling’s review of the request and result; adult NSFW requests are rejected without a generation charge. SpicyAPI adds no separate filter to Kling 2.1, and cannot disable Kling’s checks. MiniMax H3 Spicy is a separate unrestricted choice from Kling 2.1, with its own input options. Before submitting to Kling 2.1, obtain the rights and consent for the source material and take responsibility for the resulting video.

Related models and next tasks