Kling 4.0 Focuses on Multilingual Audio and a Flash Mode

Key Takeaways

1Kling 4.0 introduces stereo audio, multilingual lip-syncing, and multiple references for video creation

2A Flash variant allows for faster and cheaper renders, limited to 720p and 20 seconds

3HDR outputs and a two-minute extension have been announced, but Kling still lags behind Wan 3.0 and Seedance 2.5 on certain criteria

Why it matters : Kling 4.0 expands creative possibilities and accessibility, but competition remains strong in terms of output quality.

The new version of Kling's video generator adds stereo audio, expanded lip-sync capabilities for major languages, and multiple references to guide staging. A Flash variant aims for faster and cheaper renders, with limits set at 720p and 20 seconds. HDR outputs and an extension of clips to two minutes are announced as upcoming, while a recent comparison with Wan 3.0 and Seedance 2.5 still places Kling behind in shot logic, physics, and dialogue. API pricing has not yet been published.

Mixed Comparison and Upcoming Features

A video compares Kling 4.0 to Wan 3.0 and Seedance 2.5, two Chinese video generators. In this comparison, Kling does not take the lead, particularly in shot consistency, physics, and dialogue quality. This situation could signal a slowdown in the pace of advancements in AI video models, pending the next versions of Wan and Seedance. Kling AI also announces the upcoming release of HDR 10-bit output in 4K and 1080p, as well as an extension of clip duration to two minutes. API pricing has not yet been communicated, with support also indicated as forthcoming.

Rich References and Reusing Existing Videos

Kling 4.0's Omni Reference feature supports up to 15 references per generation, including a maximum of 10 images and 5 videos for a total duration of 30 seconds, along with 7 additional references announced. Users also have the option to use storyboards, mockups, image mosaics, videos in white model, or using a depth map as reference materials. The Creative Recreation feature allows the use of an existing video as a base to replace certain products or characters. According to Kling AI, it is possible to individually modify facial expressions, body movements, camera perspectives, or backgrounds without affecting the rest of the sequence.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

Stereo Audio, Lip-Sync, and Multilingual Text

Kling 4.0 adds a two-channel stereo audio rendering. According to the company, lip movement alignment is optimized in more than ten languages, including German, English, Japanese, Korean, and Spanish, while taking into account different dialects and accents. Text overlay in the video is now possible in multiple languages and remains stable even during camera movements.

Cadences, Limits, and Tools: 30s, Flash 720p, MCP, and CLI

Kling 4.0 allows the creation of videos up to 30 seconds in a single operation, with the ability to use up to ten keyframes to control transitions, character postures, and key narrative steps. The number of instructions analyzed can reach a maximum of 8,000 tokens. The lighter Flash variant aims for faster and cheaper generation, with a resolution limited to 720p and a maximum duration of 20 seconds. The creation interface has been redesigned, and Kling 4.0 is accessible via MCP and command line (CLI), allowing for workflow automation with AI agents. Kling AI claims that the Flash variant can reduce the gap between AI-generated videos and professional productions. Version 4.0 also brings improved resolution and more precise controls.