Kling 2.6 is Now Available on fal
We are pleased to announce the release of Kling 2.6, now available exclusively on fal on day 0. This latest model in the Kling family introduces native audio generation for both text-to-video and image-to-video generation, expanding creative possibilities across a wide range of use cases.
Kling 2.6's Top Use Cases
Cinematic Videos
Kling 2.6 shines in emotionally driven cinematic storytelling, delivering both high quality visuals and remarkably expressive audio performances. In the earthquake-rescue example, the model not only recreates the dust-choked atmosphere and gritty realism of the scene but also nails the strain, urgency, and raw humanity in the rescue worker’s hoarse voice.
Visual: The ruins after an earthquake, with twisted steel bars and concrete blocks interwoven, dust filling the air, and a rescue dog wandering nearby. Dialog: [Rescue worker] kneels on a pile of rubble, straining to pry open a broken concrete slab, then begins digging through the debris with his bare hands, his gloves bleeding at the fingertips. [Rescue worker, hoarse voice] shouts loudly into a narrow gap: "Stay with me! Can you hear me? Your daughter wants you to know... her drawings guided us right to you!"
Fast cinematic arc shot. The camera rapidly orbits 180 degrees around the freezing woman, starting from her front profile and ending behind her shoulder. The background trees whip by with motion blur (parallax effect). The woman is kneeling in the snow, hugging herself. The camera maintains a consistent distance from her. High consistent texture, volumetric lighting.
Visual and Sound Effects
For high-intensity VFX, Kling 2.6 introduces a leap in both detail and sound design. In the mechanical bomb video example below, the model nails visual and sound effects such as the crackle of ignition, the deep bass of the expanding fireball, and the sound of the fire. This makes Kling 2.6 a standout choice for creators working on action sequences, VFX previz, and .
Macro probe lens shot. The camera moves inside the complex brass gears of a ticking mechanical bomb timer. We see microscopic dust and oil on the gears. Suddenly, the spark ignites. The camera pulls back rapidly (reverse dolly) as the mechanism explodes. We witness the explosion in slow motion, expanding from a tiny spark to a massive fireball that engulfs the room. Debris flies past the lens. Extreme detail, fire simulation.
Product Ads
For product-focused storytelling, Kling 2.6 excels at clean environments, controlled camera motion, and polished dialogue delivery. In the fashion livestream example, the host’s enthusiastic delivery feels natural and well-timed, capturing the upbeat rhythm of real promotional speech. In the robotic vacuum demo, the narration blends seamlessly with ambient cleaning sounds, enhancing the sense of authenticity. Across both examples, Kling 2.6 demonstrates strong consistency, precise lip-sync, and a commercial-ready level of clarity
Visual: In a fashion live-streaming room, clothes hang on a rack, and a full-length mirror reflects the host's figure. Dialog: [African-American female host] turns to show off the sweatshirt fit. [African-American female host, cheerful voice] says: "360-degree flawless cut, slimming and flattering." Immediately, [African-American female host] moves closer to the camera. [African-American female host, lively voice] says: "Double-sided brushed fleece, 30 dollars off with purchase now.
Visual: In a tidy living room, a white robotic vacuum sits in the center, with no clutter around it. Dialog: [Narator, soft female voice] accompanied by the gentle sound of vacuuming: "Are you still troubled by dust in hard-to-reach corners? This robotic vacuum features edge-to-edge cleaning, leaving no gaps behind—making your life easier and effortless!" The camera closely follows the vacuum's path as it cleans.
Social Media Videos
Finally, the native audio generation makes the model great for funny AI social media content. The prompt adherence is strong and the voices generated very realistic, which allows creators to make great content.
Dog firefighter rescues kittens from a tree
Endpoints
Audio Prompt Guide
Multicharacter Dialogue Prompt Examples and Guidelines
| Principle | Guideline | Correct Example | Incorrect Example |
|---|---|---|---|
| P1. Structured Naming | Character labels must be unique and consistent. Avoid pronouns or synonyms. | [Character A: Black-suited Agent] and [Character B: Female Assistant] |
[Agent] says... Then, he says... |
| P2. Visual Anchoring | Bind dialogue to a character’s unique actions. Describe the action first, then the dialogue. | The black-suited agent slams his hand on the table.[Black-suited Agent, angrily shouting]: “Where is the truth?” | [Black-suited Agent]: “Where is the truth?” (Model won’t know who slammed the table) |
| P3. Audio Details | Assign unique tone and emotion labels to each character. | [Black-suited Agent, raspy, deep voice]: “Don’t move.”[Female Assistant, clear, fearful voice]: “I’m scared.” | “[Man] says…” “[Woman] says…” (Voice descriptions too vague) |
| P4. Temporal Control | Use clear linking words to control sequence and rhythm. Optionally insert: “this is when the speaker switches.” | [Black-suited Agent]: “Why?” Immediately, [Female Assistant]: “Because it’s time.” | [Black-suited Agent]: “Why?” [Female Assistant]: “Because it’s time.” (Model may merge speech) |
Common Audio Trigger Words
| Audio Type | Category | Trigger Words | Examples |
|---|---|---|---|
| Speech | Core Speech | Speaking / Talking | A woman is sitting at a desk, calmly speaking into a microphone. |
| Asking / Querying | A curious boy in the garden asking his father a question. | ||
| Telling / Narrating | An old man sitting by the fireplace, slowly telling a story. | ||
| Explaining | A tour guide pointing at a map, clearly explaining the route. | ||
| Volume/Clarity | Whispering | Two friends leaning in close in a crowded room, whispering a secret. | |
| Softly Speaking | A student in the quiet library is softly speaking on the phone. | ||
| Clearly Speaking / Crisp Voice | A radio announcer with a clear voice is speaking the news. | ||
| Emotion/Tone | Excitedly Speaking | The award winner is holding a trophy, excitedly speaking their acceptance speech. | |
| Complaining | A customer at the counter complaining about poor service. | ||
| Sighing | A tired worker sitting by a window, letting out a heavy sighing sound. | ||
| Gently Speaking | A mother is rocking a baby, gently speaking a lullaby. | ||
| Dialogue | Interaction | Answering / Responding | The interviewee is answering the question immediately. |
| Arguing / Quarrelling | A couple in the kitchen, arguing loudly. | ||
| Shouting / Yelling | A father standing at the door is shouting/yelling at his children playing outside. | ||
| Discussing | A group of students gathered around a table, discussing a difficult problem. | ||
| Vocal Action | Action | Crying / Sobbing | A little girl sitting on the ground crying after falling down. |
| Screaming | A woman seeing a mouse, letting out a sharp screaming sound. | ||
| Laughing / Chuckling | Three people sharing a joke and laughing loudly. | ||
| Singing | Core Form | A Capella | A singer on an empty stage performs the first line a capella. |
| Humming | A chef happily humming a tune while cooking. | ||
| Loud Singing | A rock musician singing loudly from the mountaintop. | ||
| Technique/Style | Bel Canto / Opera | A soprano in a gown performing a bel canto / opera piece. | |
| Pop Vocals | A young artist in a studio recording a pop track. | ||
| Vibrato | A singer adding a beautiful vibrato to the high note. | ||
| Falsetto | A male vocalist using falsetto to hit a very high note. | ||
| Harmony / Layered Vocals | A quartet performing a section with perfect harmony. | ||
| Rap | Terminology | Rapping / Hip-Hop | A street performer rapping under neon lights. |
| Flow / Rhyme | A rapper performing a verse with smooth flow and tight rhyme. | ||
| Fast Rap / Rapid Delivery | A high-speed, machine-gun-like fast rap delivery. | ||
| Strong Rhythm / Heavy Beat | A hip-hop track with a strong rhythm and heavy beat. | ||
| SFX | Daily Actions | Tapping / Knocking | A carpenter tapping a nail with a hammer. |
| Footsteps | Slow and heavy footsteps walking in an empty hallway. | ||
| Chewing / Munching | A person munching on crunchy chips. | ||
| Material Impact | Glass Shattering | A rock hits a window, followed by glass shattering. | |
| Metal Clanging | Two large iron blocks clanging in a factory. | ||
| Friction / Rubbing | The sound of rough fabric rubbing together. | ||
| Natural Elements | Thunder | A flash of lightning followed by thunder. | |
| Fire Crackling | A campfire burning and crackling. | ||
| Bubbling / Gurgling | Hot soup bubbling on the stove. | ||
| Mechanical Noise | Alarm / Siren | A police car passing by with its siren wailing. | |
| Braking | A car screeching to a stop. | ||
| Gears Whirring | The inner gears of a clock quietly whirring. | ||
| Musical Instruments | Piano Music | A pianist playing classical piano in a concert hall. | |
| Guitar Plucking | A street artist gently plucking a guitar string. | ||
| Ambient | Urban | Traffic Noise / Car Flow | Continuous traffic noise at a busy intersection. |
| Crowd Murmur | Background sound of crowd murmur in a museum. | ||
| Subway Noise | A train arriving and departing at a station. | ||
| Construction Noise | Persistent construction noise in the city. | ||
| Nature | Ocean Waves | The soothing sound of ocean waves in the morning. | |
| Bird Chirping | Birds chirping in a morning forest. | ||
| Wind Sound | Wind blowing across an open field. | ||
| Rainforest | A rainforest filled with bird calls and dripping water. | ||
| Indoor Space | Library Silence | Deep library silence with the occasional book drop. | |
| Café Background Music | Quiet café music with soft chatter. | ||
| Air Conditioner Hum | The steady hum of an air conditioner. | ||
| Fireplace Burning | The warm crackling sound of a fireplace. |
Stay tuned to our Youtube, Reddit, blog, Twitter, or Discord for the latest updates on generative media and the new model releases!