Kling 4.0, Kuaishou’s latest AI video model, is now fully available to the public as of Thursday, October 15, 2026 — and it matters because AI video is shifting from short, one-shot clips toward directed scenes creators can actually plan, revise, and finish.
October 15, 2026 — Kling AI confirmed the full release in an announcement today, ending the limited early-access period that began in late September.
What happened
Kling AI, the video-generation unit of Chinese tech company Kuaishou, formally released Kling 4.0 on October 15. The model is available through a refreshed web creation experience, and the launch also adds MCP and CLI access, so teams can plug generation into their own tools and workflows. Kling 4.0 was first unveiled on September 28, when a lighter version called Kling 4.0 Flash opened to a limited group of users for early access.
The details that matter
The headline capability is control. A single generation runs from 3 to 30 seconds — double the previous generation’s maximum — at resolutions up to 4K. Creators can place up to 10 keyframes in one generation to choreograph how a scene unfolds, instead of describing the whole clip in one prompt and hoping for the best. A feature called Omni Reference accepts up to 15 reference items in total, including up to 10 images and 5 reference videos, to keep characters, products, and lighting consistent across shots. Audio is generated in stereo with lip-sync across nine languages, and prompts can run up to 8,000 tokens.
The faster Flash variant makes 3-to-20-second clips at 720p for quick drafts. Two pieces of the September announcement are still listed as coming soon rather than available today: 10-bit HDR output at 1080p and 4K, and enterprise and developer API access. Video extension up to roughly two minutes, stitched across several generations, is also still on the roadmap.
Why it matters
Kling is not a side project. Kuaishou has been preparing the unit as a standalone business with a Hong Kong listing as the end goal, and every model milestone doubles as a pitch to investors. For working creators, the practical story is the workflow change: keyframes and reference budgets attack the continuity problems — faces drifting, products changing shape, rooms rearranging — that have kept AI clips hard to deliver for clients. Drafting in Flash and finishing in the full model is now a realistic production pattern.
The honest caveat is to separate what shipped today from what is still promised. The web, MCP, and CLI entry points are live; the broadcast-grade HDR output and the API that automated pipelines would need are not yet. Budget a trial before quoting a client deadline on either.
Frequently asked questions
Is Kling 4.0 available to everyone now?
Yes. Kling announced the full release on October 15, 2026, through its web creation experience, with MCP and CLI access also available. The earlier Flash version had been limited to a small early-access group since September 28.
How long a video can Kling 4.0 generate?
One generation runs 3 to 30 seconds, up to 4K resolution. Extension to about two minutes across multiple generations is listed as coming soon.
What is still missing from the launch?
10-bit HDR output at 1080p and 4K, plus enterprise and developer API access, are both listed as coming soon rather than available at the full release.
Sources: Kling AI announcement (via DEV Community coverage), Monolith, TechTimes, CryptoBriefing