Kling 2.1 Master
What is Kling 2.1 Master?
Kling 2.1 Master is the Pro-only variant of Kling 2.1. It always renders at 1080p and supports both text-to-video and image-to-video from one model ID. Pass a prompt on its own for a fresh generation, or pass a prompt plus a start frame to animate a supplied still. Duration is 5 or 10 seconds and a sound effects track is available in the same call. There is no Standard toggle and no end-frame parameter, which is the entire point of Master: one input shape, one output resolution, predictable downstream QC.
Key features of Kling 2.1 Master
Four features cover the case for Master over the rest of the 2.1 line.
Pro-only 1080p output
Every render is 1080p. There is no Standard mode to forget about, which keeps QC pipelines simple and removes the resolution drift you get from a per-job toggle.
Text-to-video and image-to-video in one model
Send a prompt on its own for a fresh generation, or pass a start frame alongside the prompt to animate a supplied still. The same model ID handles both paths.
5 or 10 second clips
Pick the length that fits the placement. 5 seconds for feed loops and short ads, 10 seconds for product reveals and longer narrative beats.
Sound effects in the same call
The sound_effects object adds ambient sound, background music, or ASMR-style audio to the result. Omit the field for a silent clip.
Best for
Hero spots that need 1080p
Always-Pro output for ads and brand placements without a per-job toggle. Useful when the team should never deliver 720p by accident.
Premium social posts
1080p in vertical or square. The right resolution for paid social and feed-grid presence.
Pitch and review delivery
High-resolution exports without exposing lower-quality variants to stakeholders during review.
Brand-safe single-profile pipelines
One Pro path simplifies QC. Useful for agencies and in-house teams with strict review.
Product teasers from a single still
Optional start image plus prompt for controlled motion in 1080p.
Single-quality production at scale
When the entire team should output the same resolution, Master enforces it at the model layer.
Use cases
Run a campaign at a single resolution by routing every brief through Master, so the QC step only ever evaluates 1080p output. Generate a 10-second product teaser from key art by passing the still as the start frame and a short motion prompt. Layer in a music bed through sound_effects to skip the separate audio pass. Build a paid social cut at 9:16 by uploading a vertical hero still and running Master against it; the output inherits the input aspect, so the asset is feed-ready without a reframe.
API examples
Call Kling 2.1 Master from any language by POSTing to /v1/tasks. Full parameter docs live at docs.unifically.com/models/video/kling/kling-2.1.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "kuaishou/kling-2.1-master-video",
"input": {
"prompt": "Cinematic scene with dramatic lighting",
"duration": 10,
"sound_effects": {
"music": "cinematic orchestral"
}
}
}'
Successful submission returns a task_id. Poll GET /v1/tasks/<task_id> or set a callback_url on the request to receive the finished result.
FAQs
People also ask
Kling 2.1 Master is the Pro-only variant of the 2.1 line. It always renders at 1080p, supports text-to-video with an optional starting image, and accepts an optional sound effects track. Duration is 5 or 10 seconds. There is no Standard toggle and no end-frame parameter.
Regular Kling 2.1 requires a start frame and lets you pick Standard or Pro. Master locks the output to Pro (1080p) and supports both text-to-video (prompt only) and image-to-video (prompt plus optional start frame).
Move up when you need end frames in the text path (added in 2.6), native audio with voice references (2.6), 15-second clips, multi-shot storytelling (3.0), or 4K output (3.0). Master stays useful when you want a single 1080p output from the 2.1 family with no toggles.
Master is one path with one quality profile. Useful for QC pipelines that should only ever evaluate Pro 1080p output and cannot risk a Standard render slipping through. 2.6 has more features but more variance in what comes back.
Output inherits the input image aspect when a start frame is present. For pure text-to-video the model returns 16:9 by default.