Kling

Motion and creativity: Kling brings Kuaishou's latest video models into a unified workflow. From text or image to video, motion control with reference assets, AI avatars driven by audio, to 3.0 multi-shot storytelling—each mode is tuned for quality, prompt adherence, and fluid motion.

Multi-mode first: Nine distinct modes cover text-to-video, image-to-video, motion control, talking-head avatars, and advanced 3.0 single/multi-shot. Choose the right mode for your project without leaving the platform.

v2.5 Turbo

Image-to-video Pro: Single image (optional tail frame), duration 5s or 10s, negative prompt and CFG scale. Prompt up to 2500 characters.

Text-to-video Pro: Aspect ratio 16:9, 9:16, 1:1; duration 5s or 10s; negative prompt and CFG scale for fine control.

2.6 series

Text to Video: Sound on/off, aspect ratio 1:1, 16:9, 9:16; duration 5s or 10s.

Image to Video: Single image input, sound, duration 5s or 10s.

Motion Control: Reference image plus reference video; character orientation from image or video; output 720p or 1080p.

AI Avatar

Standard and Pro: Upload avatar image and audio; generate talking-head video. Prompt up to 5000 characters. Pro tier for higher quality output.

3.0 Video

Single shot or multi-shot; image references; duration 3–15 seconds; optional sound; standard or pro mode.

Social and short-form: Create text-to-video or image-to-video clips for TikTok, Reels, and Shorts with flexible aspect ratio and 5–10s duration.

Motion control: Use reference image and video to steer character pose and motion for consistent, controllable results.

AI avatars: Turn avatar image and voiceover into talking-head videos for explainers, dubbing, and personalized content.

Multi-shot storytelling: Use 3.0 Video with multiple image references and 3–15s duration for scenes and narrative clips.

Advertising and marketing: Animate product shots and concept art with v2.5 Turbo or 2.6; add motion control or avatars as needed.

Prompt length: Up to 2500 characters (v2.5 Turbo); up to 5000 characters (AI Avatar, others as per mode).

Input: Image(s) JPEG/PNG/WebP; video for motion control; audio for AI Avatar. Size and format limits follow each mode.

Duration: 5s or 10s (v2.5, 2.6); 3–15s (3.0 Video).

Resolution: 720p or 1080p where applicable (e.g. motion control, 2.6); aspect ratios 1:1, 16:9, 9:16 (and 21:9 for some modes).

Output: Video via URL or download; optional sound where supported.

Choose mode: Pick the Kling mode that matches your goal—v2.5 Turbo I2V/T2V, 2.6 text/image/motion, AI Avatar Standard/Pro, or 3.0 Video.

Upload assets: Provide image(s), and for motion control add reference video; for AI Avatar add audio.

Set parameters: Select duration, aspect ratio, resolution, and options (sound, negative prompt, CFG, etc.) as shown in the form.

Generate: Submit and wait for the result; preview in the result panel and download when ready.

Prompt crafting: Be specific about motion, style, and composition. For avatars, clear speech and consistent tone in the audio improve lip-sync and expression.

Mode choice: Use v2.5 Turbo for fast, high-quality text/image-to-video; 2.6 for sound and motion control; AI Avatar for talking-head; 3.0 for multi-shot and longer clips.

Resolution and duration: Match duration to platform (e.g. 5–10s for shorts). Use 1080p when the mode supports it for final deliverables.

Kling on FuseAITools brings Kuaishou's latest video models to your workflow. From text or image to video, motion control, AI avatars, and 3.0 multi-shot—choose the right mode and create video in seconds.