Kling 3.0 is Now Available on fal

We're pleased to announce the release of Kling 3.0, now available on fal from day zero. Kling 3.0 represents a new state-of-the-art generation stack for video and image creation, built for structured storytelling rather than isolated clips.

Team fal

• 5 min read
Kling 3.0 is Now Available on fal

We're pleased to announce the release of Kling 3.0, now available on fal from day zero. Kling 3.0 represents a new state-of-the-art generation stack for video and image creation, built for structured storytelling rather than isolated clips. It supports storyboarding with up to six shots, consistent characters through reusable elements, and richer multi-modal prompting that can incorporate both audio and image inputs. The result is a model family designed for creators who want more control, continuity, and cinematic direction in their workflows.

0:00
/0:08

Playground link: https://fal.ai/models/fal-ai/kling-video/o3/pro/text-to-video/playground?share=f61a3de6-150c-4966-9d58-da14e34b2e28

At the center of the release is Kling 3.0, the new core model for both text-to-video and image-to-video generation. The model supports clips from 3 to 15 seconds in duration, along with start- and end-frame conditioning. Alongside 3.0, Kling Video o3 represents the most advanced tier of the omni-video lineup, designed for higher-end customization and storyboard-first creation.

Key Strengths

With Kling 3.0 and O3, Kling supports:

  • Character consistency
  • Voice binding, which means that characters can maintain consistent voices across generations
  • Up to six shots, each with its own prompt and duration, with total clip lengths of up to 15 seconds.
  • Reference conditioning, including support for video references

This is a major shift: instead of describing an entire scene in one paragraph, you can now direct it shot-by-shot.

0:00
/0:10

This video was generated using Multi-shot prompting. The prompt included 2 scenes for the seperate cuts that merge to produce the final result.

Model Improvements

Kling 3.0 comes with major qualitative upgrades that go beyond resolution or longer clips. The focus is clearly on making generated video feel more believable, more directed, and more usable in real creative workflows. Across acting, voice, motion, and editing, the new models showcase state-of-the-art capabilites.

Realistic Acting

One of the most noticeable improvements in Kling 3.0 is the jump in character performance. Facial motion is significantly more natural, dialogue pacing feels better timed, and gestures carry more realism. Instead of stiff or robotic movement, characters show more convincing acting beats; subtle expressions, smoother body language, and stronger continuity across shots.

0:00
/0:08

Playground link: https://fal.ai/models/fal-ai/kling-video/o3/pro/text-to-video/playground?share=03c253ad-462e-4c61-82d7-1f38ff0209ad

0:00
/0:05

Playground link: https://fal.ai/models/fal-ai/kling-video/o3/pro/text-to-video?share=9f7ff5b3-82af-4672-97a2-b4112f4b68a2

Advanced Voice Control

Kling 3.0 also strengthens the connection between voice and character. Voice-to-subject matching is more consistent, with improved tonality and more natural dialogue speed. Spoken lines feel less synthetic and more grounded in the scene, making character-driven clips much more compelling. A great example is the “mother cooking” style demo, where voice delivery and facial performance align in a way that feels closer to real video than previous generations.

0:00
/0:10

Playground link: https://fal.ai/models/fal-ai/kling-video/o3/pro/text-to-video/playground?share=e2d60474-3fc5-4243-b52a-63cd9269ed30

0:00
/0:08

Playground link: https://fal.ai/models/fal-ai/kling-video/o3/pro/text-to-video/playground?share=fee10cc3-ee3e-47bc-92fb-f9ff8a9baa08

0:00
/0:08

Playground link: https://fal.ai/models/fal-ai/kling-video/o3/pro/text-to-video/playground?share=e321a947-0ded-4b4e-8994-7337dfcb8ef5

Motion Control

Fast-paced motion has historically been one of the hardest challenges for generative video, often introducing warping or visual artifacts. Kling 3.0 shows a clear improvement here, producing significantly cleaner results even when scenes involve rapid movement, action, or camera shifts.

0:00
/0:08

Playground link: https://fal.ai/models/fal-ai/kling-video/o3/pro/text-to-video/playground?share=3a49532d-8670-4fc2-b49b-5d37bb381ef6

Video Editing

Finally, Kling 3.0 expands editing-oriented generation compared to earlier models like O1. With stronger reference video support, creators can rely on Kling as an editor, changing backgrounds, modifying clothing, inserting or removing people, and reshaping scenes while preserving the original structure. These capabilities unlock entirely new workflows, from AI-assisted post-production to remixing existing footage into new cinematic variations.

0:00
/0:08

Change the background to an interrogation room

Image Generation & Editing

Kling 3.0 also brings meaningful upgrades on the image side with Kling Image 3.0 The new image models support sharper outputs up to 4K resolution, stronger prompt adherence, and improved consistency when working with faces or reusable elements. Alongside text-to-image, Kling Image 3.0 enables more powerful image-to-image editing workflows, making it easier to refine style, modify subjects, or generate coherent visual series that pair naturally with Kling’s video storytelling stack.

Before/After Image Slider — Recraft
Input image Recraft output (after)
Input
Recraft

Endpoints

Kling O3 pro :

Kling O3 standard :

Kling 3.0 pro

Kling 3.0 standard

Kling 3.0 Image

Stay tuned to our Youtube, Reddit, blog, Twitter, or Discord for the latest updates on generative media and the new model releases!