Kling AI¶

Development line: project:kling-ai · thread kling-ai-product-development
Last event: 2025-06-30 · 2 dated since 2025-06-30 · Researched: 2026-09-04 · confidence: medium

What it is¶

Kling AI is Kuaishou's web and app platform for creating short commercial, narrative, and character videos.

• VIDEO 3.0 generates text-to-video and image-to-video with native audio, multi-shot sequences, and dialogue.
• VIDEO 3.0 Omni applies image, video, and element references to maintain consistent characters or products.
• Motion Control and Lip Sync transfer motion or add dialogue to finished human characters.

VIDEO 3.0 generates 3–15 seconds per pass; Lip Sync requires a fully visible face and excludes animals. Use Kling for short, controlled scenes with references and sound; assemble long videos from verified shots.

Development line¶

• 2025-06-30 — Kling AI exposed a dedicated video-to-lip workflow. On 2025-06-30, a dated link targeted Kling AI's dedicated video-to-lip workflow page. This establishes that the product surfaced video-to-lip as a distinct workflow at that point. The link alone does not establish its launch date, feature details, availability, or regional access.
• 2025-06-30 — Kling AI exposed a dedicated video-to-audio workflow. On 2025-06-30, a dated link targeted Kling AI's dedicated video-to-audio workflow page. This establishes that the product surfaced video-to-audio as a distinct workflow at that point. The link alone does not establish its launch date, feature details, availability, or regional access.

What changed¶

2024-07-07 — The web version launched Kling publicly with start/end frames, shot control, and text-to-video up to 10 seconds.
2024-09-20 — Kling AI 1.5 added 1080p, Motion Brush, and Lip Sync.
2024-11-20 — We found no confirmed Kling release on this date and leave the model line unchanged.
2025-06-30 — Video-to-Audio turned audio generation into a standalone route; Video-to-Lip routed to existing Lip Sync rather than a new model.
2026-02-04 — Kling 3.0 unified Video 3.0, Video 3.0 Omni, Image 3.0, and Image 3.0 Omni, adding up to 15 seconds, native audio, references, and multi-shot generation.

How to use this¶

As of 2025-06-30, practitioners should evaluate Kling AI's video-to-lip and video-to-audio workflows separately when selecting a production path; their detailed capabilities and availability still require source-level verification.

1. For a new scene, open VIDEO 3.0, pick text-to-video or image-to-video, set duration to 3–15 seconds, and enable Native Audio when sound should generate with picture. — https://kling.ai/quickstart/klingai-video-3-model-user-guide
2. For a recurring character, product, or shot sequence, use VIDEO 3.0 Omni: build an Element from 2–4 images or a short video and use it as a reference. — https://kling.ai/quickstart/klingai-video-3-omni-model-user-guide
3. Describe the scene by setting, subject, action, sound, style or emotion, and camera; place dialogue in quotes and link it to the speaking character. — https://kling.ai/quickstart/klingai-video-26-audio-user-guide
4. For existing footage of a person, select Lip Sync, upload or generate speech, and check results before publishing. — https://kling.ai/quickstart/ai-lip-sync-guide
5. For directed motion, upload a movement video and character image, attach a facial Element if needed, and verify orientation alignment. — https://kling.ai/quickstart/motion-control-user-guide

Best practices¶

• Write prompts as a director's brief: scene, subject, motion, sound, and camera, not a pile of adjectives. — https://kling.ai/quickstart/klingai-video-26-audio-user-guide
• Use multiple angles or a short reference video for a consistent face; a single photo forces the model to guess unseen sides. — https://kling.ai/quickstart/motion-control-user-guide
• Name which character speaks each line to prevent mixed voices in dialogue. — https://kling.ai/quickstart/klingai-video-3-model-user-guide
• Use only owned or licensed material for Voice Control; supply 5–30 seconds from a single speaker with even delivery and minimal background noise. — https://kling.ai/quickstart/klingai-video-26-audio-user-guide
• Provide video with a fully visible human face for Lip Sync, and fit audio length to the clip duration. — https://kling.ai/quickstart/ai-lip-sync-guide

Superseded by this¶

• 2024-07-07 — The 10-second limit is obsolete for VIDEO 3.0: the current guide specifies 3–15 seconds, native audio, and multi-shot.
• 2024-09-20 — Kling 1.5 is no longer the active working version; Motion Brush and early Lip Sync remain legacy features in the wider 3.0 family.
• 2025-06-30 — New sound scenes no longer require silent video followed by separate audio: VIDEO 2.6 and 3.0 generate picture, speech, effects, and ambience in one pass. Standalone Lip Sync remains the route for redubbing finished clips.
• 2026-02-04 — The Ultra-only early access notice reflected a temporary release state; check current tier and regional access inside your account.

Still unknown¶

• The 2024-11-20 entry cannot be classified: the direct link returned no text, and no matching primary release was found.
• Several original app.klingai.com release note pages were not machine-readable; we cross-checked their facts against official Kuaishou releases and Kling guides.
• Current subscription tiers, credits, regional availability, and access to 3.0 Turbo, CLI, and MCP were not verified through an authenticated account.
• Available primary sources indicate Kling and Kling AI refer to the same Kuaishou product; no distinct entity under the short name Kling is confirmed.

Sources¶

Agent brief¶

• Subject: project:kling-ai, thread kling-ai-product-development, 2 dated events 2025-06-30 → 2025-06-30.
• Practical note: As of 2025-06-30, practitioners should evaluate Kling AI's video-to-lip and video-to-audio workflows separately when selecting a production path; their detailed capabilities and availability still require source-level verification.
• Confidence: medium. Dated supersedes above are the authority for what is obsolete.