Kling 3.0's 15-second cinematic fantasy clip surprises with character consistency
eyishazyer · x · 2026-09-14
The author used Kling 3.0, Kuaishou's video model, to generate a 15-second high-budget-looking dark fantasy sequence: a lone torch-carrying traveler inside a vast underground cavern.
He shared the full prompt, with key techniques:
- Locking character appearance (hair, clothing, build, props) in detail and demanding full consistency across the sequence;
- Specifying environment and atmosphere (massive stone formations, mysterious carvings, thick mist) for cinematic quality.
His prior experience with AI video: shots stutter, lighting goes flat under camera movement, and the illusion drifts apart after a few seconds. Kling 3.0's character consistency and camera stability surprised him.
Related event: Kling 3.0 tested: stable camera moves and strong character consistency(2 posts)→
More from Multimodal
- Gemini 4 Carbon vs Opus 5.5 hands-on: 3x cheaper at $4.10 and twice as fast — vitaliychiley · 2026-10-31
- Surreal AI fairytale: a town obsessed with turning tears into yarn — gizakdag · 2026-10-31
- ComfyUI extension adds one-click saving of preview images with metadata — sitefall · 2026-10-31
- DeepMind scientist directs MKUltra microdrama with Gemini agents for ~$35 total — korymath · 2026-10-31
- Using a gamepad-controlled Godot camera path to drive Wan video generation — droidsculptor · 2026-10-31
- Customuse launches Decals to fix 3D AI models' blurry logos and text — SimplyAnnisa · 2026-10-31