The biggest jump in AI video this season is not sharper pictures — it is longer, talking clips. Kuaishou's Kling 4.0 is now fully live for everyone on klingai.com, after a September preview, and it generates complete 30-second videos — with the soundtrack created in the same pass.
What is actually new
- 30 seconds in one generation. Kling 3.0 topped out at 15 seconds. Thirty seconds is a full product shot, a complete social-media hook, or a short scene with a beginning, middle and end — no stitching clips together.
- Sound is generated with the picture. Every clip comes with native two-channel stereo audio: ambient sound, music and dialogue, with lip sync for talking characters across nine languages, including English, Spanish, German, French and Hindi.
- Omni Reference. Feed it up to 15 reference assets — faces, products, settings — and it holds them consistent across the scene, the feature advertisers and product-video makers will care about most.
- Up to 10 keyframes. Pin the start, middle and end of a shot and let the model fill in the motion between them, for far more control than a single text prompt.
Rollout has been quick: within the last two weeks Kling 4.0 has also arrived inside partner platforms creators already use, including Runway and Google Flow, so you may already have access without a new account.
The honest notes
Two things the launch headlines skip. First, the 4K and 10-bit HDR output Kling has announced is still listed as "coming soon" — today's model works at up to 1080p through its API partners. Second, quality still varies from clip to clip: the first generation is a draft, and the keyframes and references exist precisely because steering takes a few tries. Budget your credits for retries.
How to try it free, step by step
- Go to klingai.com — always the official site, because copycat sites appear around every big AI launch — and sign in. New accounts receive free credits (your first clips carry a watermark).
- Choose text-to-video, or image-to-video if you have a product photo or a starting frame.
- Write the scene like a director: subject, action, setting, camera movement. Add reference images if a face or product must stay consistent.
- For a talking clip, write the dialogue and pick the language — the lip sync and voice are generated with the video.
- Generate, review, and re-roll the weak shots with keyframes before spending credits on a longer final cut.
Why creators should care
AI video used to mean five silent seconds and a lot of editing. A 30-second clip with dialogue, sound and a consistent product in frame is a different product entirely: it is most of a short ad, made from a laptop. The gap between "AI demo" and "usable content" is closing fastest exactly here.
Sources: Pandaily (Kling 4.0 full launch on klingai.com, 16 October 2026), TechSifted, and AI Tool Steps' Kling 4.0 guide; features cross-checked against Kling's official site on 4 November 2026.
