Kling released the video model Kling V 2.6, enabling direct 1080p video generation with integrated audio and multilingual lip-sync.

Core functionality

The update supports both text-to-video and image-to-video pipelines, yet does not currently support composition from two keyframes in production workflows.

Users may specify exact phrases in prompts for synchronized speech, and the model can generate accurate lip movements across multiple languages.

The system also accommodates non-human subjects, enabling video generation with animals while preserving visual and audio coherence within produced clips.

Prompts accept explicit textual instructions that specify spoken phrases and timing, enabling synchronized dialogue generation and fine-grained control over delivered audio.

Availability

Kling V 2.6 is already distributed through several aggregators, including Freepik, Fal, and Higgsfield, for developers and creators among other partners.

Access methods and licensing terms vary by aggregator, and usage limits or attribution requirements depend on each platform's policies.

O1 Image announcement

Yesterday Kling introduced O1 Image, an image generator positioned as an alternative to Banana, though it has been available since the O1 Video launch.

The company did not publish additional technical specifications or commercial details at the presentation, leaving integration specifics to partners and aggregators.

Developers and creative teams can test Kling V 2.6 through supported aggregators, following each platform's onboarding and licensing procedures.