Kling 4.0 goes global: 30-second AI video, 15 reference assets, and a pitch at the pros

• Kling AI, the video platform of Hong Kong-listed Kuaishou Technology, announced the global launch of Kling 4.0 on October 16, 2026.
• The model generates up to 30 seconds of video in a single pass, double the previous generation's 15 seconds.
• Omni Reference supports up to 15 reference assets (images, video, audio) and up to 10 keyframe inputs for frame-level control.
• The three stated priorities: realism, creative control, and professional-grade output.
• Kling AI reports more than 100 million users globally and 50,000+ enterprise customers across advertising, film, TV, and gaming.
• Coming soon: 10-bit HDR output at 4K and 1080p, plus a 21:9 ultrawide format.

Kuaishou's Kling AI launched Kling 4.0 globally on October 16, 2026 — the next generation of its AI video model, promising steadier motion, native 30-second clips, and fine-grained control for professional creators. It's a direct bid to take AI video out of the one-off-clip era and onto real production timelines.

What's new in 4.0

The launch notes center on three upgrades. First, steadier dynamic motion and complex camera movements — the kind of shots (fast action, orbiting cameras) where AI video tools have historically melted faces and drifted objects mid-clip. Second, native 30-second generation in a single pass, doubling the previous generation's 15-second maximum and letting a character move, react, and shift emotion in one continuous take. Third, Omni Reference: up to 15 reference assets (images, video, audio) and up to 10 keyframe inputs, for frame-level control over characters, scene transitions, and story progression.

Aimed at the pros

The positioning is explicit: Kling AI says it's working with professional creators to deliver the key shots that define a work's quality, not just the occasional viral clip. The company reports more than 100 million users globally and over 50,000 enterprise customers across advertising, film, television, and gaming. On the roadmap: 10-bit HDR output at both 4K and 1080p plus a 21:9 ultrawide format — specs that speak to colorists and editors, not casual prompters.

Reading the fine print

Independent analysis by AI Parabellum adds texture to the announcement: reference capacity jumps from seven to 15 per job, but the multi-reference and keyframe modes can't be mixed in a single task, and audio uploads serve as voice references rather than music tracks. The same report notes the launch notes mention MCP support and a command-line tool — a hint that Kling is courting automated workflows and AI agents, not just browser-based creators. Kuaishou's own announcement, dated October 16, says the full model was already available to the public from October 15.

Why it matters

The AI video race has been stuck at short, unstable clips; the contenders that break that ceiling with controllable, production-friendly output win the budgets that actually pay for this technology. Kling 4.0's pitch — longer native takes, character consistency via references, and professional output formats — reads like a playbook for winning enterprise accounts. Whether the steadier motion holds up in real productions is the question every filmmaker will now be testing.