von KuaishouVideo

Kling

27 Modelle · ab 425 Credits

Modelle der Familie Kling

Modell	Modus	Preis	Seitenverhältnisse
Kling Effects	Video	ab 425 Credits	—	Öffnen →
Kling v2.5 Turbo Pro Image-to-video	Bild-zu-Video	ab 595 Credits	—	Öffnen →
Kling v2.5 Turbo Pro Text-to-video	Text-zu-Video	ab 595 Credits	—	Öffnen →
Kling v2.6 Pro Avatar	Sprechender Avatar	Hängt von deiner Datei ab	—	Öffnen →
Kling v2.6 Pro Image-to-Video	Bild-zu-Video	ab 1.190 Credits	—	Öffnen →
Kling v2.6 Pro Motion Control	Bewegungssteuerung	Hängt von deiner Datei ab	—	Öffnen →
Kling v2.6 Pro Text-to-Video	Text-zu-Video	ab 1.190 Credits	—	Öffnen →
Kling v2.6 Std Avatar	Sprechender Avatar	Hängt von deiner Datei ab	—	Öffnen →
Kling v2.6 Std Motion Control	Bewegungssteuerung	Hängt von deiner Datei ab	—	Öffnen →
Kling v3.0 Pro Image-to-Video	Bild-zu-Video	ab 1.428 Credits	—	Öffnen →
Kling v3.0 Pro Motion Control	Bewegungssteuerung	Hängt von deiner Datei ab	—	Öffnen →
Kling v3.0 Pro Text-to-Video	Text-zu-Video	ab 1.428 Credits	—	Öffnen →
Kling v3.0 Std Image-to-Video	Bild-zu-Video	ab 1.071 Credits	—	Öffnen →
Kling v3.0 Std Motion Control	Bewegungssteuerung	Hängt von deiner Datei ab	—	Öffnen →
Kling v3.0 Std Text-to-Video	Text-zu-Video	ab 1.071 Credits	—	Öffnen →
Kling Video 4.0 Image-to-VideoNeu	Bild-zu-Video	ab 2.139 Credits	—	Öffnen →
Kling Video 4.0 Reference-to-VideoNeu	Referenz-zu-Video	ab 2.139 Credits	16:9 · 9:16 · 1:1 · 21:9	Öffnen →
Kling Video 4.0 Text-to-VideoNeu	Text-zu-Video	ab 2.139 Credits	16:9 · 9:16 · 1:1 · 21:9	Öffnen →
Kling Video 4.0 Video-EditNeu	Videobearbeitung	Hängt von deiner Datei ab	—	Öffnen →
Kling Video O3 Pro Image-to-Video	Bild-zu-Video	ab 1.190 Credits	—	Öffnen →
Kling Video O3 Pro Reference-to-Video	Referenz-zu-Video	ab 952 Credits	—	Öffnen →
Kling Video O3 Pro Text-to-Video	Text-zu-Video	ab 1.190 Credits	—	Öffnen →
Kling Video O3 Pro Video-Edit	Videobearbeitung	Hängt von deiner Datei ab	—	Öffnen →
Kling Video O3 Std Image-to-Video	Bild-zu-Video	ab 952 Credits	—	Öffnen →
Kling Video O3 Std Reference-to-Video	Referenz-zu-Video	ab 714 Credits	—	Öffnen →
Kling Video O3 Std Text-to-Video	Text-zu-Video	ab 952 Credits	—	Öffnen →
Kling Video O3 Std Video-Edit	Videobearbeitung	Hängt von deiner Datei ab	—	Öffnen →

Regeln der LUVI Prompt Engine für Kling 3

TARGET MODEL: Kling 3 (Kuaishou video — v3.0 Pro cinematic / Omni O3 native lip-synced audio). It reads CINEMATIC INTENT, not inventories: direct a scene for a cinematographer, never list objects.

Regeln, die gelten müssen

Every shot is the formula, left-to-right: Subject + Subject Movement + Scene + (Camera Language + Lighting + Atmosphere). Who, doing what, where, then how it's shot and lit. A muddy clip is a missing slot, not a bad seed.
Concise beats maximal: 60-100 words, 2-5 sentences per shot, one job per sentence. Never an adjective wall. HARD CAPS: the whole prompt at most 2,500 characters (overrun fails the submission); each shot card ~512 characters — if it won't fit in 512, it's two shots.
Kill garnish: delete "cinematic, 4K, dramatic, trending". At most 1-2 earned mood words as a tail AFTER the direction is written (serene, intimate, epic).
ONE clean camera move per shot, named by its grip term — "slow push-in", "pull-back", "smooth 180° orbit", "handheld with subtle shake", "tracking shot". "Push in then pan right" never works: two moves means two shots. Name no move and the shot drifts — pin it.
Camera pace rides the camera verb: lead camera moves with slow / smooth / steady. Action intensity rides the adjective on the ACTION verb, matched to the user's intent ("slow deliberate wave" vs "frantic rapid wave"). Kill dead verbs ("moves", "goes") — load the verb with speed + direction + posture + contact point ("races down a rain-soaked street, weaving between cars").
ONE action per shot — if the user's prompt carries more beats, split into a multi-shot board rather than stacking actions in one shot. Stage the action in chronological beats with an ENDPOINT — preparation, main action, follow-through, landed ("sets it down. Soft clink when cup meets saucer."). Open-ended action drifts.
Name the light fixture and its direction, never the feeling: "neon signs", "candlelight", "rim light from top-right", "soft key 45° camera-left" — never "good/nice/dramatic lighting".
Exclusions: leave unwanted content OUT of the prompt entirely. Never write "no X" in the prompt body — unwanted things are simply not described.

Was am besten funktioniert

Anchor motion in physics and a described environment — weight, contact, cloth response. Motion in a void distorts; motion that touches a real world stays coherent.
Time + weather are lighting presets: "golden hour", "night", "morning mist", "full moon". Atmosphere = particles that make light visible ("dust in the air", "haze", "drifting smoke") with a hard source raked through.
Grade in words at the tail: temperature + contrast + format — "warm tones, high contrast, 35mm film grain".
Style is a named medium as a tail clause, never "artistic": "Studio Ghibli animation", "film noir lighting and shadows", "1980s VHS camcorder aesthetic", "photorealistic rendering", "cel-shaded anime".
Ambient sound lives in the scene slot ("rain tapping on the roof") — audio generates in the same pass; contact words double as SFX.
Dialogue (native audio): action FIRST, then the line — stage the physical beat so the model knows whose mouth to move: "The agent slams his hand on the table. [Black-suited Agent, angrily shouting]: \"Where is the truth?\"". Label speakers uniquely and re-cite the exact label every time (a synonym reads as a new person); tag delivery inside the bracket; sequence turns with "Immediately," and hold beats with "Pause." / "Silence."; lay ambience first ("[Sound: low electronic hum and falling rain.]"); match word count to clip length — a 5s shot holds one or two short lines.
Multi-shot boards: one clean beat per shot card, each line ordered framing → subject → subject's motion → camera move; connect cuts with "Continuing…"/"then" (match cut), "Immediately" (hard cut), "Camera cuts to…" (explicit); repeat load-bearing subject descriptors word-for-word across cuts ("man in red suit"); roughly 6 shots ceiling, ~512 characters per shot, whole board within the 2,500-character cap.
Character consistency: one unique unchanging label per character, cited every beat — kill "he"/"the man"/synonyms; keep the identity block byte-identical when it recurs.

Aus den Regeln der LUVI Prompt Engine für diese Familie; dieselben Regeln wenden der Rewriter im Arbeitsbereich und LUVIs MCP-Connector an.

Regeln der LUVI Prompt Engine für Kling 4.0

TARGET MODEL: Kling 4.0 (Kling AI / Kuaishou — native multimodal video with stereo audio generated in the same pass, 3–30 s). It acts as a director: a short prompt is expanded with its own storyline, blocking, camera and even a line of dialogue; a clear prompt is followed faithfully. Multi-shot is ON by default with no switch. No negative_prompt, no seed, no cfg, no weights — the prose is the whole instrument.

Regeln, die gelten müssen

DESCRIBE WHAT THE CAMERA SEES. Kling's own principle: "Use descriptive language and minimize abstract terms and negative instructions." Abstract labels ("premium look", "vintage feel", "cinematic", "cinematic feel", "dynamic camera", "high quality", "masterpiece", "8k") do nothing useful: when the user's own words already give the concrete detail behind a label, keep the detail and delete the label; when they do not, delete the label — never invent the detail yourself. Never add such labels.
A WRITTEN TIME RANGE IS A SHOT. "0–3 s: … 3–7 s: … 7 s–end: …" produces a hard cut exactly at each boundary, even when the range describes a continuous move. Use ranges ONLY when the user wants separate shots or beats that must land at given times. To keep one continuous take across several beats, write them as one flowing sentence in order and add "One continuous shot, no cuts." You do NOT know the clip length (it is a parameter you cannot see): keep the user's own time marks exactly, and never invent second marks — when the user lists shots without times, keep them as ordered shots ("Shot 1: … Shot 2: …"), never convert them to seconds.
PROSE THAT MOVES FROM PLACE TO PLACE INVITES CUTS. Several actions in several locations without ranges come back as a montage. If the user wants one take, keep the action continuous and the camera following it.
KEEP A RECURRING PERSON CONSISTENT: use the user's description of a person once, word for word, and refer to them by the same label in every later shot (never a synonym). Kling changes an undescribed person's face and age across cuts — but do not invent a description; if the user gave none, keep the user's label. When the user names a count, add a count guard in Kling's form: "EXACTLY 2 CHARACTERS — NO DUPLICATES."
AUDIO IS ALWAYS GENERATED. Write the soundtrack the scene needs: the user's dialogue/voiceover + ambient sounds and effects that the described place and actions make. NEVER add music, a song or a score the user did not ask for. When subtitles or background music are not wanted, end with "No subtitles. No background music." — these are the two exclusions Kling says are always worth writing.
DIALOGUE: visible speaker + action first, then "speaks in <language>, <manner>:" and the line in double quotes, in that language's own script: She speaks in natural, clear, slightly slow Spanish: "¿Dónde está la estación?" Describe age, tone, speed and emotion for a specific voice. One speaker per turn, no overlaps, short lines for the seconds available; time several turns with ranges. Keep the user's line word for word.
ON-SCREEN TEXT goes in double quotes, short, ideally one headline line; keep the user's copy character for character, diacritics included. Subtitles the user DOES want: state position, number of lines, style and timing.
EXCLUSIONS: describe the scene positively ("He is shown in profile", not "don't show his face") — one positive choice, never "from behind or in profile". Keep only exclusions Kling itself prescribes: subtitles/background music when unwanted, what a reference must not carry over, "no photorealism" for animation styles, or one specific thing the brief would otherwise invite. Never stack generic negative lists.
DURATION, RESOLUTION AND ASPECT ARE PARAMETERS. Never write "16:9", "4K", "1080p" or "15 seconds long" into the prompt as settings. Ranges inside the prompt are fine.
No weights, no "(term:1.3)", no "--ar", no "seed:", no "Negative prompt:" label — delete them, keep the words they wrapped.

Was am besten funktioniert

Kling's template, blocks dropped when they do not apply: [Overall Settings] Scene and texture: …; visual style: …; camera tone: … · [Sequence of Events] 0–X s: … · [Audio] … — or plain prose for a single take.
Two registers. A ONE-LINE IDEA ("An old man ranting about AI.") is a deliberate request for Kling's own direction: return it essentially as written — fix grammar, delete dead labels, at most add "No subtitles. No background music." — and never write a plot, a dialogue line, wardrobe or a room around it. A BRIEF the user already structured (headers like SCENE CONTEXT, ACTION TIMING, AUDIO, or time ranges) is already in Kling's form: keep its header labels verbatim ("SCENE CONTEXT:" stays "SCENE CONTEXT:"), its ranges exactly as written ("9 s–end" stays "9 s–end"), its wording and order; change only what breaks a MUST rule. Longer is not better.
Camera: when the user names a move, write it with what it follows ("the camera tracks beside her at her pace"); a vague one ("dynamic camera") becomes the one move the user's action implies (following the moving subject) or is deleted. One dominant move per shot. Never invent amounts or percentages.
Beat density: about three clean beats in 10 s, two beats plus one line in 8 s; a 5 s clip with three actions becomes a cut-up montage. For 20–30 s keep motion slow to moderate and give the final seconds a written action.
When the user names a drink or food, name its vessel too if the user's words imply it ("sips her Turkish coffee" → "sips from a small Turkish-coffee cup") — Kling otherwise picks a generic cup.

Aus den Regeln der LUVI Prompt Engine für diese Familie; dieselben Regeln wenden der Rewriter im Arbeitsbereich und LUVIs MCP-Connector an.

Regeln der LUVI Prompt Engine für Kling 4.0 Video Edit

Regeln, die gelten müssen

DESCRIBE WHAT THE CAMERA SEES. Kling's own principle: "Use descriptive language and minimize abstract terms and negative instructions." Abstract labels ("premium look", "vintage feel", "cinematic", "cinematic feel", "dynamic camera", "high quality", "masterpiece", "8k") do nothing useful: when the user's own words already give the concrete detail behind a label, keep the detail and delete the label; when they do not, delete the label — never invent the detail yourself. Never add such labels.
A WRITTEN TIME RANGE IS A SHOT. "0–3 s: … 3–7 s: … 7 s–end: …" produces a hard cut exactly at each boundary, even when the range describes a continuous move. Use ranges ONLY when the user wants separate shots or beats that must land at given times. To keep one continuous take across several beats, write them as one flowing sentence in order and add "One continuous shot, no cuts." You do NOT know the clip length (it is a parameter you cannot see): keep the user's own time marks exactly, and never invent second marks — when the user lists shots without times, keep them as ordered shots ("Shot 1: … Shot 2: …"), never convert them to seconds.
PROSE THAT MOVES FROM PLACE TO PLACE INVITES CUTS. Several actions in several locations without ranges come back as a montage. If the user wants one take, keep the action continuous and the camera following it.
KEEP A RECURRING PERSON CONSISTENT: use the user's description of a person once, word for word, and refer to them by the same label in every later shot (never a synonym). Kling changes an undescribed person's face and age across cuts — but do not invent a description; if the user gave none, keep the user's label. When the user names a count, add a count guard in Kling's form: "EXACTLY 2 CHARACTERS — NO DUPLICATES."
AUDIO IS ALWAYS GENERATED. Write the soundtrack the scene needs: the user's dialogue/voiceover + ambient sounds and effects that the described place and actions make. NEVER add music, a song or a score the user did not ask for. When subtitles or background music are not wanted, end with "No subtitles. No background music." — these are the two exclusions Kling says are always worth writing.
DIALOGUE: visible speaker + action first, then "speaks in <language>, <manner>:" and the line in double quotes, in that language's own script: She speaks in natural, clear, slightly slow Spanish: "¿Dónde está la estación?" Describe age, tone, speed and emotion for a specific voice. One speaker per turn, no overlaps, short lines for the seconds available; time several turns with ranges. Keep the user's line word for word.
ON-SCREEN TEXT goes in double quotes, short, ideally one headline line; keep the user's copy character for character, diacritics included. Subtitles the user DOES want: state position, number of lines, style and timing.
EXCLUSIONS: describe the scene positively ("He is shown in profile", not "don't show his face") — one positive choice, never "from behind or in profile". Keep only exclusions Kling itself prescribes: subtitles/background music when unwanted, what a reference must not carry over, "no photorealism" for animation styles, or one specific thing the brief would otherwise invite. Never stack generic negative lists.
DURATION, RESOLUTION AND ASPECT ARE PARAMETERS. Never write "16:9", "4K", "1080p" or "15 seconds long" into the prompt as settings. Ranges inside the prompt are fine.
No weights, no "(term:1.3)", no "--ar", no "seed:", no "Negative prompt:" label — delete them, keep the words they wrapped.

Was am besten funktioniert

Kling's template, blocks dropped when they do not apply: [Overall Settings] Scene and texture: …; visual style: …; camera tone: … · [Sequence of Events] 0–X s: … · [Audio] … — or plain prose for a single take.
Two registers. A ONE-LINE IDEA ("An old man ranting about AI.") is a deliberate request for Kling's own direction: return it essentially as written — fix grammar, delete dead labels, at most add "No subtitles. No background music." — and never write a plot, a dialogue line, wardrobe or a room around it. A BRIEF the user already structured (headers like SCENE CONTEXT, ACTION TIMING, AUDIO, or time ranges) is already in Kling's form: keep its header labels verbatim ("SCENE CONTEXT:" stays "SCENE CONTEXT:"), its ranges exactly as written ("9 s–end" stays "9 s–end"), its wording and order; change only what breaks a MUST rule. Longer is not better.
Camera: when the user names a move, write it with what it follows ("the camera tracks beside her at her pace"); a vague one ("dynamic camera") becomes the one move the user's action implies (following the moving subject) or is deleted. One dominant move per shot. Never invent amounts or percentages.
Beat density: about three clean beats in 10 s, two beats plus one line in 8 s; a 5 s clip with three actions becomes a cut-up montage. For 20–30 s keep motion slow to moderate and give the final seconds a written action.
When the user names a drink or food, name its vessel too if the user's words imply it ("sips her Turkish coffee" → "sips from a small Turkish-coffee cup") — Kling otherwise picks a generic cup.

Aus den Regeln der LUVI Prompt Engine für diese Familie; dieselben Regeln wenden der Rewriter im Arbeitsbereich und LUVIs MCP-Connector an.

Regeln der LUVI Prompt Engine für Kling Motion Control

TARGET MODEL: Kling Motion Control (Kuaishou, v2.6 and v3.0). Motion transfer: a character IMAGE plus a driving VIDEO; the character performs the clip's movement and expressions. The image supplies the character, background and every other element of the picture. The clip supplies the action. The prompt is a side channel for what neither shows — it is optional, and an empty prompt is a correct answer.

Regeln, die gelten müssen

Never describe the action. The clip IS the choreography; Kling's own guide says the action need not be described, and over-directing movement fights the reference motion. Delete every sentence that narrates what the body does (spins, steps, raises arms, dances, walks, turns, jumps) — even when the user wrote it.
If nothing remains after the choreography is removed, return an EMPTY prompt. Do not pad it, do not restate the task ("replace the character in the video with the character from the image" is inert), do not write a placeholder.
Never contradict the image. Face, hair, body, costume and set come from the image; a prompt that changes them fights the picture. Wardrobe detail is allowed only where it agrees with what the user said is in the image.
Write only what the pixels cannot carry: wardrobe detail, an added background subject ("a corgi runs in, circling around her feet"), light, hour, weather, atmosphere, style ("neon lighting, cyberpunk city, volumetric fog"; "golden hour, film grain").
Camera: never ADD camera words. A camera move written by the user is kept only if it is one of Kling's presets — zoom in, zoom out, camera up, camera down, fixed position — and it takes effect only when the character follows the image; when it follows the clip, the clip's own camera is used.
Plain prose only, at most 2,500 characters. No weight syntax, no brackets, no dialogue lines, no shot cards, no timestamps — the model generates no speech and no lip-sync.
Delete resolution and quality garnish ("4K", "masterpiece", "ultra detailed"), durations, aspect ratios and negative lists. Length comes from the clip, framing from the image or the clip, and there is no negative prompt.

Was am besten funktioniert

One or two short sentences. Most good motion-control prompts are a clause of light and setting, or nothing.
Keep the user's own nouns for anything they want added to the scene.

Aus den Regeln der LUVI Prompt Engine für diese Familie; dieselben Regeln wenden der Rewriter im Arbeitsbereich und LUVIs MCP-Connector an.