Models / Kling / Text to video
Kling text to video prompting guide
Last checked 2026-10-11Kling 3.0 / O3 (Kuaishou); 2.x legacy · Kling 3.0 (V3), O3; 2.5 Turbo Pro t2v only; Kling 4.0 announced 2026-09-28
How to prompt Kling in text-to-video mode. Every point links its source; community tips are marked. Your agent gets the same guide through the Atlas MCP (get_guide).
Prompt structure
subjects anchor→scene→action over time→camera→audio→style
- Write prompts as directions to a scene, not object lists. Multi-shot (up to 6 shots): label each shot with framing, subject and motion. source ↗
What the text should describe
- Anchor subjects early; describe what the audience should see and feel as the scene unfolds. source ↗
- T2V template: Setting → Subject → 'Action: First [state], then [main action], finally [resolution]' → Camera (move + lens) → Style → Audio. Prompt length ~100–200 words for VIDEO 3.0 (40–60 for 2.5 Turbo). community source ↗
- Single-take multi-beat progression with 'First … Next, without a cut, … Finally …' and a continuity line ('rigid brush geometry throughout'). community source ↗
- Kling 4.0 Flash: shorter, straightforward prompts (Kling 3.0 style) worked better for the creator than long Seedance-style prompts. community source ↗
Camera vocabulary
- Shot language: profile shot, macro close-up, tracking shot, POV, shot-reverse-shot. Moves: slow push-in, dolly left, crane up, handheld, locked-off, orbit; lens feel 35mm/85mm. community source ↗
- Describe camera behaviour over time: tracking, following, freezing when the subject pauses, panning, moving in sync. source ↗
Responds well to
- Define core subjects at the start and keep their descriptions identical across shots. source ↗
Duration and limits
- Kling V3/O3 on fal: 3–15 s, up to 1080p; aspect 16:9, 9:16, 1:1. source ↗
- Longer durations risk drift (fal skill). community source ↗
Multi-shot
- multi_prompt list or shot_type 'intelligent'/'customize'; either prompt or multi_prompt, not both. source ↗
Avoid and failure modes
- Prefer direct declarative prompts; skip prestige adjectives; when negative_prompt is exposed list specific things, not vague ones. community source ↗
Negative prompt
Supported via negative_prompt (default: “blur, distort, and low quality”). source ↗
Real prompts, credited
they deserved a higher score Kling 4.0 Flash prompt 👇 Realistic vertical handheld phone footage at a bright indoor ice rink during a figure skating competition, no text or captions on screen. Two enormous sumo wrestlers in mawashi loincloths and top-knots perform a graceful pairs figure skating routine: gliding hand in hand, one lifts the other high overhead with one arm and spins, then throws him into a triple jump that lands with a thud that cracks the ice, both finishing with elegant arms out while judges in the background hold up score cards. Single continuous shot, no cuts, romantic orchestral music, scraping blades and a heavy thud.
Test 8: Multi-Beat Narrative Continuity & Temporal Persistence A key architectural leap in Kling 4.0 is handling single-take narrative progressions. This benchmark evaluates whether the model remembers objects and actions across distinct sequential beats within one continuous camera move. Prompt: Continuous medium-tracking sequence in a sunlit art studio. First, an artist dips a wooden paintbrush into a glass jar of deep blue watercolor, causing ink plumes to bloom in the water. Next, without a cut, the artist lifts the wet brush and applies a single smooth stroke onto a clean canvas. Finally, the artist sets the brush down on a wooden easel ledge and steps back slightly. Consistent natural lighting, seamless chronological progression, rigid brush geometry throughout. Duration: 10s
Master Prompt: Joker begins his iconic dance descent down the stairs… Multi shot Prompt 1: … (Duration: 5 seconds) Multi shot Prompt 2: … (Duration: 5 seconds)
Next
All Kling modesKling model, prices & shotsFrom $0.07/s · pricesSeedance text to videoVeo text to videoRunway text to videoLuma text to video
Sources
- blog.fal.ai/kling-3-0-prompting-guide/
- fal.ai/kling-3
- fal.ai/models/fal-ai/kling-video/o3/pro/reference-to-video/api
- fal.ai/models/fal-ai/kling-video/v3/pro/image-to-video/api
- github.com/fal-ai-community/skills/blob/HEAD/skills/fal-prompting/references/kling.md
- github.com/maciejdzierzek/kling-ai-prompt-generator/blob/main/README.md
- x.com/Artedeingenio/status/2104829034299351079
- x.com/sebatheepan/status/2104653416123551799