curl-X POST 'https://pollo.ai/api/platform/v1/generation/kling-ai/kling-v2-1/video'\-H'Content-Type: application/json'\-H'x-api-key: YOUR_API_KEY'\-d'{ "input": { "prompt": "A cinematic shot of a golden retriever running through a field of sunflowers at sunset, warm rim light, shallow depth of field", "negativePrompt": "blurry, low quality, distorted, watermark, text", "image": "https://example.com/first-frame.jpg", "duration": 5, "mode": "std", "strength": 50 }}'
Optionally include webhookUrl to receive a callback when the task succeeds or fails.
Optional request field
{"input":{"prompt":"A cinematic shot of a golden retriever running through a field of sunflowers at sunset, warm rim light, shallow depth of field","negativePrompt":"blurry, low quality, distorted, watermark, text","image":"https://example.com/first-frame.jpg","duration":5,"mode":"std","strength":50},"webhookUrl":"https://example.com/webhooks/pollo"}
Schema
Input fields for the selected endpoint and mode.
Field
Type
Required
Description
prompt
string
No
Text prompt describing the content to generate.
negativePrompt
string
No
Text describing elements to avoid in the generated output.
image
string
Yes
First-frame / reference image URL (HTTP(S).
duration
number
No
The duration of the generated video, in seconds.
mode
string
No
Generation mode of the model (e.g. a faster or higher-quality tier).
strength
number
No
Prompt adherence strength; higher values follow the prompt more strictly.
Kling 2.1 is an upgraded video model with improved prompt adherence, steadier character motion, and sharper visuals. It supports text-to-video and image-to-video generation.
Through Pollo API, developers add Kling 2.1 text-to-video and image-to-video to their products, choosing standard, pro, or 4K output and 5 or 10 second clips per request.
Key Features of Kling 2.1 API
Faithful Prompt Following
Described subjects, actions, and staging land closely on screen, so a written scene translates into the intended shot instead of a loose approximation that needs reworking.
Steadier Character Motion
People and creatures move with more natural weight and continuity, keeping limbs, posture, and gait coherent across the clip rather than warping during quick or complex actions.
Start and End Frame Control
Supply a first frame and a tail frame, and Kling 2.1 fills the motion between them, giving precise control over how a shot begins and resolves.
Tiered Output Quality
Pick standard for fast drafts, pro for cleaner detail, or 4K when a shot needs to hold up on large screens, matching effort to the delivery format.
Prompt Strength and Negatives
A strength control adjusts how firmly the model obeys the prompt, while negative prompts steer unwanted elements out, giving directable results without endless retries.
Use Cases of Kling 2.1 API
Keyframe Transitions: Set a beginning and ending image and let Kling 2.1 generate the in-between motion for controlled reveals, morphs, and product state changes.
Character Motion Clips: Animate a described figure walking, gesturing, or reacting where consistent body movement matters more than heavy visual effects.
Still Photo Animation: Feed a single image and bring it to life as a short moving shot while keeping the original framing and subject intact.
Explainer Snippets: Produce short 4K-ready clips for tools that need clean, prompt-accurate motion inside tutorials and how-to sequences.
Concept Previz: Let creative teams see a described shot before committing to a full production, trusting output to match the written direction.
Social Video Automation: Generate 5 or 10 second vertical or wide clips at volume for content apps that batch many prompts through one pipeline.
Iterative Drafting: Use standard mode to test many variations quickly, then re-render the winner in pro or 4K for the final asset.
How to Use Kling 2.1 API
Get an API Key: Create a Pollo API account and generate your key from the developer dashboard.
Choose the Model: Send requests to the Kling 2.1 endpoint at /generation/kling-ai/kling-v2-1.
Add Your Inputs: Provide an image for image-to-video plus a prompt, optionally add imageTail for end-frame control, then set mode (std, pro, or 4k) and length (5 or 10).
Generate and Retrieve: Submit the task, poll the returned task status, and download the finished video once processing succeeds.
Prompting Best Practices for Kling 2.1 API
Because Kling 2.1 rewards precise direction, write prompts that name the subject, the exact action, the camera behavior, and the setting, then trim unwanted elements with a negative prompt.
A Simple Prompt Formula
Subject + specific action and pacing + camera movement + setting and lighting + style, with negatives for anything to exclude
What You Should Notice
Be Literal About Action: State exactly what moves and how, since the model follows described motion closely rather than inventing its own.
Use Negative Prompts: List elements to avoid, such as extra hands or text overlays, to clean up results without re-rolling.
Plan Keyframes: When using start and tail frames, make both images share lighting and framing so the generated motion stays smooth.
Tune Prompt Strength: Raise strength for strict adherence, lower it when you want the model more freedom to interpret the scene.
Match Mode To Purpose: Draft in standard, finalize in pro or 4K so you spend render time only on the shots you keep.
Example Prompts
Keyframe Reveal
"A closed origami crane on a wooden table slowly unfolds itself flat into a single sheet of paper. Overhead camera, soft daylight from a nearby window, calm minimal style. Start frame folded, tail frame flat."
Character Motion
"A young violinist draws a long bow stroke, shoulders rising with the phrase, eyes closing in concentration. Slow dolly-in, warm rehearsal-room light, realistic style. Negative prompt: extra arms, blurred hands."
Everyday Craft Shot
"A barista pours steamed milk into espresso, a leaf pattern forming on the surface as the cup fills. Tight top-down view, soft cafe lighting, clean realistic style, steady hands and smooth pour."
Nature Close-Up
"A dew-covered spider web trembles as a light breeze passes, droplets shivering but holding on. Macro shot, early morning backlight, shallow focus, naturalistic style, gentle and continuous motion."
Kling 2.1 vs Runway Gen-4 vs Luma Ray 2
Capability
Kling 2.1
Runway Gen-4
Luma Ray 2
Prompt adherence
✅ Strong, literal following
✅ Good, more interpretive
✅ Good
Character motion stability
✅ Steady limbs and gait
✅ Solid
✅ Smooth but softer
Start and end frame control
✅ First and tail frame
✅ Keyframe support
✅ Start and end frames
Resolution options
Up to 4K
Up to 1080P
Up to 1080P
Clip length
5 or 10 seconds
Short clips
Short clips
Recommended For
Prompt-faithful, high-res motion
Stylized VFX shots
Fast fluid image animation
Why Choose Kling 2.1 API?
Kling 2.1 fits products that need motion to match the brief, from keyframe transitions to steady character clips, with output that scales from quick drafts up to 4K.
Through Pollo API, you reach Kling 2.1 with one key alongside 300+ leading video and image models, with clean docs and task status polling.
Start with the Kling 2.1 endpoint on Pollo API to ship prompt-faithful, keyframe-controlled video, then compare it against Runway, Luma, and Hailuo without touching your integration.
Kling 2.1 API FAQs
What is Kling 2.1?
Kling 2.1 is an upgraded Kling AI video model focused on tighter prompt adherence, steadier character motion, and sharper visuals, supporting text-to-video and image-to-video generation.
Does Kling 2.1 API support image inputs?
Yes. You can animate a source image, and optionally add a tail image so the model generates the motion between your chosen first and last frames.
What resolutions and lengths are available?
Kling 2.1 offers standard, pro, and 4K output modes, with clip lengths of 5 or 10 seconds selected per request.
How does it handle character movement?
Its upgrade centers on continuity, so limbs, posture, and gait stay coherent through quick or complex actions instead of warping mid-clip.
Can I control how strictly it follows the prompt?
Yes. A prompt strength setting adjusts adherence, and negative prompts let you exclude unwanted elements for cleaner, more directed results.
Why run Kling 2.1 through Pollo API?
Pollo API gives one integration for Kling 2.1 and 300+ other models, with documentation and task tracking.
