Do not rewrite the frame; write how clay changes
The frame locks hands, bowl and light while the prompt owns rotation, pressure, droplets and protected structure.
Update your browser, then reload this page. Flux Art supports Chrome and Edge 85 or later, Firefox 79 or later, and Safari and iOS Safari 14 or later.
Reload Flux ArtCheck your network connection and reload this page. If the problem continues, clear your browser cache and try again.
Reload Flux ArtAI E-commerce product suites are now live
Upload a product image, choose a platform and the images you need, then create listing-ready main images, white-background images, selling-point images, and lifestyle scenes. Product suites currently support Taobao/Tmall, JD.com, Pinduoduo, and Douyin E-commerce, with every result saved together for easy review, refinement, and export.
Begin with one frame for credible motion, or lock the ending with two boundary images in Pro mode.
Kling V2.1 is an image-to-video-only route in Flux Art. The frame supplies existing facts; the prompt describes action, force and camera without redefining subject, composition or material.
The frame locks hands, bowl and light while the prompt owns rotation, pressure, droplets and protected structure.
Base and summit images own the result; the prompt owns speed, rail path, shadows and deceleration.
Define wind, scarf path and footwork to avoid intersections or unexplained imbalance.
The input route defines existing visual facts. The prompt should supply motion and change instead of competing with the source.
Add motion to one established image in Standard or Pro.
Control the ending with two boundary states in Pro mode.
Use it when a reliable first frame already exists and needs cost-effective product, human, material or environmental motion, optionally with a locked ending.
Preserve subject, proportions, composition, light and material while describing only observable action and camera.
Use two images to lock start and finish while the prompt owns path, speed, physics and continuity.
Clay, fabric, plants and mechanisms need visible force and result to avoid unexplained deformation.
Use Standard for quick validation and Pro for important motion quality or boundary-frame results.
Do not redescribe the image; spend prompt budget on action order, camera, force and protected details.
The first frame owns hands, bowl, wheel and window light. The wheel rotates steadily; hands stabilize the wall, then press once from base upward, creating continuous wet-clay rings as a few droplets move outward. Fixed close view; preserve fingers, rim, clay color and background tools.
The first frame places the blue funicular at the hill base; the last has it at the top station. Fixed wide view centered on rails; it climbs steadily, crosses two tree shadows, slows and stops. Preserve carriage, rail count, coastline and golden sun direction.
The first frame owns person, stone stairs and red scarf. A gust moves lower-left to upper-right; scarf leaves the palm, unfurls and passes before her shoulder. She steps and reaches as weight shifts. Slow lateral camera; preserve face, clothing, stairs and scarf length, with no fabric-body intersection.
Kuaishou debuted Kling 2.1 at the end of May 2025, describing broader improvements in motion, realistic physical simulation and semantic response with Standard and High Quality routes. Flux Art currently exposes first-frame video plus Pro start-and-end frames, both with negative prompts.
Create with Kling V2.1Model capabilities are summarized from public Kling AI and Kuaishou materials. Available modes, parameters and pricing follow the current Flux Art workspace.
Published
Kling V2.1 is part of the model series Kuaishou introduced at the end of May 2025 with improvements in motion, realistic physical simulation and semantic response.
No. The current Flux Art configuration exposes image-to-video only: first-frame and Pro start-and-end-frame workflows.
First-frame generation currently supports Standard and Pro modes.
The current start-and-end-frame route is Pro only.
Yes. Negative prompts are available for both current image-to-video workflows.
Describe only chronological action, camera motion, physical force and protected details; do not redescribe subjects or composition already present in the frame.
Let the images define opening and closing states, then describe the path, speed, physical response and continuity required between them.
Use Standard to validate a direction quickly, and Pro when motion quality or an accurate boundary-frame result matters more.
Commercial use depends on current Flux Art and model-provider terms and your rights to prompts and reference media. Check the latest terms before client work.