Video and Image Generation: Sora, Veo, Kling, Flux — 2026-09-29
Kuaishou's Kling AI released version 4.0 with doubled video duration (up to 30 seconds) and multi-reference support, set for October launch. Meanwhile, copyright litigation against AI video platforms intensifies: courts ruled that music generators cannot escape lawsuits without identifying specific infringing outputs, signaling tighter legal scrutiny across the sector.
Video and Image Generation: Sora, Veo, Kling, Flux — 2026-09-29
Top developments
Kling 4.0 arrives with 30-second clips and multi-reference support
Kuaishou's Kling AI announced version 4.0, doubling maximum video generation length from 15 seconds to 30 seconds and adding support for over a dozen reference types. The model is set to launch officially in October 2026, marking a significant upgrade in commercial capability.
This release directly addresses one of the competitive gaps between Chinese and Western generators. Kling 3.0 was previously ranked at Elo 1,247 on text-to-video benchmarks, maintaining strong standing across open leaderboards despite competition from Runway, Veo 3.1, and ByteDance's Seedance 2.0.
Copyright lawsuits expand: music generator Suno cannot dismiss artist claims
A U.S. federal court ruled on September 28 that Suno, an AI music generator, cannot escape copyright infringement lawsuits without explicitly identifying specific infringing outputs. Independent recording artists were not required to pinpoint exact derivative works to survive dismissal, significantly lowering the legal barrier for claimants.
The ruling establishes precedent that could accelerate similar actions against video generators. Runway AI has already faced a YouTube creator class action over alleged misuse of training data, and leaked litigation documents suggest tech executives have privately discussed concerns about "the largest theft of labor in human history" during model training.
Flux speeds up image generation; Black Forest Labs confirms FLUX 3 video in gated access
Black Forest Labs' Flux Schnell generates images in 0.5–2 seconds via API, positioning it as the fastest option for real-time applications, while Flux Pro takes 8–12 seconds. The company's FLUX 3 multimodal model—jointly training image, video, audio, and action—has entered gated early access for video; pricing has not been announced.
Local view
Chinese media coverage emphasizes competitive positioning. Ifanr's real-time review of Kling 4.0 noted that the model "topples the table again" with cinematic motion control rivaling Western tools—testing the model on complex effects (e.g., bamboo forest sword choreography) where prior versions struggled.
Huxiu reported that Kling 4.0 will launch officially in October, highlighting the aggressive timeline as Kuaishou positions itself against ByteDance's Seedance 2.0 (which ranks at Elo 1,212 on video leaderboards).
Context & numbers
Arena Rankings (as of late September 2026):
- Seedance 2.0 (ByteDance): Elo 1,212 (video generation)
- Kling 3.0: Elo 1,247 (text-to-video)
- MiniMax H3: Elo 1,357 (open-weight image-to-video)
Speed benchmarks:
- Flux Schnell: 0.5–2 seconds per image (fastest proprietary)
- Flux Pro: 8–12 seconds per image
- Midjourney: 10–30 seconds per image
Availability:
- FLUX 3 video: Gated early access only; no public release date or pricing
- Kling 4.0: October 2026 (official launch imminent)
On the radar
- Kling 4.0 October rollout: Exact launch date within October not yet specified; market impact on pricing and adoption will depend on feature rollout speed and availability tier.
- Copyright litigation wave: The Suno precedent may embolden parallel suits against Runway, Midjourney, and other platforms. Watch for class action consolidation and potential regulatory intervention in Q4 2026.
- Black Forest Labs FLUX 3 pricing: No announcement yet; timing of public video access and cost structure will be critical for adoption among professionals competing with Kling and Runway.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
Powered by
