Kling 4.0
Kling 4.0 is the video model Kuaishou's Kling AI launched worldwide on 16 October 2026, eight months after Kling 3.0. It doubles the longest single generation to 30 seconds, makes two-channel stereo sound with every clip, takes per-shot prompts, and accepts up to 15 references across images, video clips and saved characters. Kling says it also handles up to 10 keyframes and prompts of up to 8,000 tokens. fal opened API access on 23 October with four endpoints, text, image, reference and video edit, all at 720p or 1080p and priced per second by resolution. 4K, 10-bit HDR and extending a clip to two minutes are announced but not yet available.
Anvisha Pai, Co-founder & CEO, Voyager
Kuaishou · released
›What we asked for
A self-portrait of Kling 4.0, Kuaishou's video model: one continuous shot gliding through a dark film studio as rows of hanging screens light up one by one with moving scenes, the camera pushing toward a single bright screen, low stereo hum building to a chime.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
At a glance
- Price
- $0.157 a second at 720p and $0.235 at 1080p on fal, sound included, read October 23, 2026. Our five-second 720p clips cost $0.785 each.
- Length
- 3 to 30 seconds per generation, in whole seconds, twice the previous version's limit of 15
- Audio
- Always on. Every clip we made carried a two-channel track; fal lists no switch and no silent rate.
- Control
- Per-shot prompts on the text, image and reference endpoints, a first and last frame, and up to 15 references: 10 images, 5 videos and 7 saved characters, 3 of them with a voice.
- Long takes
- Our 20- and 30-second workshop prompts each came back as one unbroken take that reached every action in order, for $3.14 and $4.71 at 720p.
- Measured wait
- 145 to 224 seconds per five-second 720p clip, all nine submitted at once on fal's first day of API access. One batch, not a service guarantee.
- Coming later
- 4K, 10-bit HDR and extending a clip to two minutes are announced, not yet available.
Other Kling models: Kling 3.0 Standard, Kling 3.0 Pro. Compare all 3.
Anvisha's take
Kling 4.0 fixes the failure I most associated with Kling: it spelled the title card, both lines and the date on navy bars, where both 3.0 tiers wrote nonsense. The market finally looks like footage rather than paint, the pour keeps the hand, the kettle and the dripper, and the harbour push is steady. The new habit is cutting. Five of our nine five-second prompts came back as two or three shots: the turntable jumps to an overhead view and back, the spill is cut three ways, and the speaker cuts to a close-up for her smile. Kling 3.0 held every one of its clips as a single take. Asked for 20 and 30 seconds, though, it held one take through every beat of the workshop prompt in order, which is the job 4.0 exists for. At $0.157 a second at 720p it costs a little more than 3.0 Standard, and $0.235 at 1080p makes it the most expensive Kling short of 4K. I would use it for long takes and anything with words on screen, and keep Kling 3.0 for a five-second product shot that has to stay one shot. These are frame-based observations; every clip carries stereo sound, which I have not scored by ear.
Kling 4.0 real output: the same 9 clips every model gets
Every video model on this site gets the same 9 clips, so you can compare like with like. These are the clips Kling 4.0 returned, one try each, nothing re-rolled or retouched. Each tile shows one frame of the clip until you press play. Generated on . The 9 clips took 145 to 224 seconds each and cost $7.07 in total.
A clean product shot that holds its shape through a full slow rotation.
›What we asked for
A matte black ceramic pour-over coffee dripper rotating slowly on a white turntable, soft studio light, seamless white background, product video, no text.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
A face, lip sync and a spoken line, where the model makes audio.
›What we asked for
A woman in her 40s with short grey hair stands in a bright kitchen, looks at the camera and says 'Good morning, the coffee is ready', natural window light, handheld feel, she smiles at the end.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
Set text that stays spelled correctly while it moves.
›What we asked for
A title card reading 'MODA SUMMER SESSIONS' animates in over slow-motion footage of a festival crowd at dusk, then a smaller line 'Friday 12 June' fades in beneath it. Bold condensed type, tangerine on navy.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
A steady drone push-in with stable geometry.
›What we asked for
Drone shot pushing slowly forward over a coastal town at sunrise, terracotta roofs, fishing boats in the harbour, long shadows, cinematic and steady.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
A spill in slow motion: does water behave like water.
›What we asked for
A glass of water tips over on a wooden table and spills in slow motion, water and three ice cubes sliding across the wood, macro lens, morning light.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
Hands, a kettle and a steady pour without extra fingers or teleporting objects.
›What we asked for
Close-up of hands pouring hot water from a gooseneck kettle in slow circles into a pour-over coffee dripper, steam rising, shallow depth of field, warm cafe light.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
Many people and objects moving at once without merging.
›What we asked for
A crowded Saturday farmers market seen from slightly above, shoppers with tote bags, a dog on a leash, a busker with a guitar, bunting swaying, late morning light, gentle camera drift.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
A named style held across every frame rather than drifting to photoreal.
›What we asked for
A lighthouse on a rocky cliff in a storm, in the style of a Japanese ukiyo-e woodblock print, waves crashing in flat colour with visible wood grain, animated like moving paper cutouts.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5
returned 1280 by 720 pixels
Bring a still product photo to life without changing the product.
›What we asked for
The coffee dripper slowly fills as coffee drips through it, steam rising, the camera drifting gently around it, the dripper itself unchanged. (starting from The Nano Banana Pro product shot from the image fixture set.)
kling/v4/image-to-video
resolution: 720p, duration: 5
returned 940 by 940 pixels
How the cost under each picture was worked out: fal Kling 4.0 720p $0.157/second for 5 requested seconds (fal.ai/models/kling/v4/text-to-video and fal's pricing API, read 2026-10-23); fal does not return the charge.
Additional workflow tests
Separate from the common benchmark above. These tests examine longer shots, successive edits or specialized controls. One attempt per test, including failures. Costs are estimates from published rates.
20-second continuous workshop shot
Kling 4.0
One unbroken shot, stable apron, mug and window; the action must reach the final pose.
›What we asked for
One continuous 20-second take in a sunlit pottery workshop, no cuts. A woman wearing a plain blue apron stands at a wooden table with one red mug and a folded white cloth. First third: she picks up the red mug and turns it once in her hands. Middle third: she places it back on the table and wipes beside it with the white cloth. Final third: she folds the cloth, places it to the left of the mug, and looks toward the same window. The camera slowly tracks sideways throughout. Keep the same woman, apron, red mug, table and window. Natural workshop ambience, no music, no text.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 20
returned 1280 by 720 pixels
Generated . Cost basis: fal Kling 4.0 720p $0.157/second for 20 requested seconds (fal.ai/models/kling/v4/text-to-video and fal's pricing API, read 2026-10-23); fal does not return the charge. Original output.
Returned 20.04 seconds at 1280x720 with stereo sound, estimated $3.14, after about 331 seconds. Scene-change detection found no cut, and the sampled frames reach every beat in order: she turns the red mug, sets it down, wipes beside it, folds the cloth and looks toward the window. A second folded white cloth appears on the table while she wipes and is gone by the end, and the camera barely moves instead of tracking sideways. This is frame review, not proof of uninterrupted continuity.
30-second continuous workshop shot
Kling 4.0
One unbroken shot, stable apron, mug and window; the action must reach the final pose.
›What we asked for
One continuous 30-second take in a sunlit pottery workshop, no cuts. A woman wearing a plain blue apron stands at a wooden table with one red mug and a folded white cloth. First third: she picks up the red mug and turns it once in her hands. Middle third: she places it back on the table and wipes beside it with the white cloth. Final third: she folds the cloth, places it to the left of the mug, and looks toward the same window. The camera slowly tracks sideways throughout. Keep the same woman, apron, red mug, table and window. Natural workshop ambience, no music, no text.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 30
returned 1280 by 720 pixels
Generated . Cost basis: fal Kling 4.0 720p $0.157/second for 30 requested seconds (fal.ai/models/kling/v4/text-to-video and fal's pricing API, read 2026-10-23); fal does not return the charge. Original output.
Returned 30.04 seconds at 1280x720, estimated $4.71, after about 393 seconds. Again no cut detected, and the beats arrive in order, ending on a look toward the window. A second cloth again sits folded on the table while she wipes with another, and the folded cloth ends on the right of the mug as we see it, her left, where the 20-second clip put it on ours.
Kling 4.0 at 1080p: animated title card
Kling 4.0
The common title-card prompt at fal's default 1080p instead of the benchmark's 720p: does the extra resolution sharpen the small date line?
›What we asked for
A title card reading 'MODA SUMMER SESSIONS' animates in over slow-motion footage of a festival crowd at dusk, then a smaller line 'Friday 12 June' fades in beneath it. Bold condensed type, tangerine on navy.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 1080p, duration: 5
returned 1920 by 1080 pixels
Generated . Cost basis: fal Kling 4.0 1080p $0.235/second for 5 requested seconds (fal.ai/models/kling/v4/text-to-video and fal's pricing API, read 2026-10-23); fal does not return the charge. Original output.
Returned 1920x1080, estimated $1.175, after 150 seconds. Both lines and the date spelled again, this time as large stacked tangerine type over the crowd with no navy bars, with the same cut at about three seconds as the 720p clip. This is a new generation, not the 720p clip sharpened, and Kling 3.0 Pro garbled this card at 1080p: the spelling comes from the model, not the pixels.
Kling 4.0 at 1080p: busy scene
Kling 4.0
The common farmers-market prompt at 1080p instead of 720p: faces, hands and bunting at more pixels, for half again the price.
›What we asked for
A crowded Saturday farmers market seen from slightly above, shoppers with tote bags, a dog on a leash, a busker with a guitar, bunting swaying, late morning light, gentle camera drift.
kling/v4/text-to-video
aspect_ratio: 16:9, resolution: 1080p, duration: 5
returned 1920 by 1080 pixels
Generated . Cost basis: fal Kling 4.0 1080p $0.235/second for 5 requested seconds (fal.ai/models/kling/v4/text-to-video and fal's pricing API, read 2026-10-23); fal does not return the charge. Original output.
Returned 1920x1080, estimated $1.175. A denser market than the 720p clip, with the busker, a golden retriever on a red lead and bunting overhead, in one unbroken shot. Our connection dropped while waiting; the runner resumed the same fal request rather than sending a new one, so the recorded 213-second wait includes that gap.
Compare Kling 4.0
Kling 4.0 pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| fal | Second at 720p Sound included; there is no audio switch. Text, image and reference-to-video all charge this rate. | $0.157 | fal model page, read |
| fal | Second at 1080p 1080p is fal's default resolution. | $0.235 | fal model page, read |
| fal | Second at 720p, video edit or video references $0.141 at 1080p, billed on the source plus the output length, at least twice the output. fal's pricing API reports $0.157 for the edit endpoint; this is the model page's figure. | $0.094 | fal video-edit model page, read |
Is Kling 4.0 free?
Kling says 4.0 is fully available on its own website, where accounts get free daily credits and paid plans, but we could not find an official credit rate for 4.0. fal and other hosts charge per second with no free tier.
- •Kling app: free daily credits and paid plans, as for Kling 3.0. On 23 October 2026 we found no per-second credit rate for 4.0 published by Kling; third-party figures disagree.
- •fal and other hosts: per second, no free tier.
Where to use Kling 4.0
- Kling app · Kuaishou's own app, where 4.0 launched first.
- fal, text-to-video · Hosted API and playground; the endpoint used for eight of the clips below.
- fal, image-to-video · First frame plus an optional last frame; used for the image-to-video clip.
- fal, reference-to-video · Images, video clips and saved characters cited in the prompt. Not tested here.
- fal, video edit · Rewrites a 3 to 30 second clip from an instruction. Not tested here.
- Voyager · Runs inside Voyager, the agent for creative work, on Voyager credits on every paid plan, or with your own fal key.
Voyager
Run Kling 4.0 inside Voyager
Voyager is an agent for creative work that runs every model in this reference, Kling 4.0 included. Brief it, and it picks the model, makes the thing and hands back something you can edit. Create an account to get started.
What we noticed running Kling 4.0
- •Both lines and the date spelled correctly, tangerine on navy bars, held through a cut from the crowd to the stage. Both Kling 3.0 tiers garbled this card. See the animated title card clip.
- •The dripper keeps its shape with no handle, but the one-shot prompt came back with two cuts: side view, a view straight down into the cone, then side view again. See the product turntable clip.
- •A convincing tip and pool, cut into three shots in five seconds; it ends on two cubes standing in the puddle. See the physics clip.
- •The photographed dripper keeps its shape, but coffee runs down its outside from the first second and pools on the base while it fills like a cup. See the image to video clip.
- •A dense, natural market with the busker and a golden retriever on a red lead, without the painted finish both Kling 3.0 tiers gave the people. See the busy scene clip.
- •Five of our nine five-second clips contain a cut that the prompt did not ask for (text card, turntable, spill, speaker and woodblock). Scene-change detection found none in any of Kling 3.0's eighteen clips.
Kling 4.0 limits and API parameters
| Limit | Value |
|---|---|
| Duration | 3 to 30 seconds per generation, in whole seconds, at 720p or 1080p. fal API schema, read |
| Output size | Our 16:9 720p clips came back at 1280 by 720, 24 frames a second, 5.04 seconds long, with 44.1 kHz stereo sound; the square product photo gave a 940 by 940 clip. fal API schema, read |
| Keyframes | Kling announces up to 10 keyframes; fal's image endpoint takes a first and a last frame, and we found no wider keyframe parameter. fal API schema, read |
| Not yet available | 4K, 10-bit HDR and extending a clip to two minutes are announced as coming soon. fal Kling 4.0 page, read |
›Every setting, for developers
fal (kling/v4/text-to-video)
fal API schema, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Up to 26,000 characters Required unless multi_prompt is set. | |
| multi_prompt | array | Shots, each a prompt of up to 512 characters and a duration of 1 to 30 seconds The clip's length is the sum of the shots. | |
| duration | integer | 3 to 30 | 5 |
| resolution | enum | 720p, 1080p | 1080p |
| aspect_ratio | enum | 16:9, 9:16, 1:1, 21:9 | 16:9 |
fal (kling/v4/image-to-video)
fal API schema, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| start_image_url | URL | Image, at least 300 pixels a side, aspect ratio 0.4 to 2.5 Required. The first frame; the clip takes its shape. | |
| end_image_url | URL | Image Optional last frame. |
fal (kling/v4/reference-to-video)
fal API schema, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| image_urls | array | Up to 10, cited as @Image1, @Image2 in the prompt | |
| video_urls | array | Up to 5, 3 to 30 seconds each and 30 seconds combined, cited as @Video1 | |
| elements | array | Up to 7 characters or objects, cited as @Element1 Each a front view plus one to three more angles, or a video; up to 3 may carry a voice_id. | |
| aspect_ratio | enum | auto, 16:9, 9:16, 1:1, 21:9 auto follows the references. | 16:9 |
fal (kling/v4/video-to-video/edit)
fal API schema, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| video_url | URL | 3 to 30 seconds, at least 700 pixels a side, up to 300 MB Required; cited as @Video1 in the instruction. | |
| image_urls | array | Up to 10 reference images | |
| duration | integer | 3 to 30 Defaults to the source clip's length. |
What Kling 4.0 will not make
Kling runs under Kuaishou's content rules, which follow Chinese regulation: coverage notes it blocks politics, protests and criticism of the government alongside the usual bans. On fal the model sits behind fal's content checks. None of our nine prompts was refused, including the festival title card that every Wan endpoint declined.
Alternatives to Kling 4.0
Frequently asked questions
What is Kling 4.0?
Kuaishou's video model, launched worldwide on 16 October 2026 and on fal's API from 23 October. It makes clips of 3 to 30 seconds with stereo sound, from text, a first and last frame, up to 15 references, or an existing clip to edit.
How much does Kling 4.0 cost?
On fal, $0.157 a second at 720p and $0.235 at 1080p, sound included, read on 23 October 2026: $0.785 for a five-second 720p clip and $7.05 for 30 seconds at 1080p. Video edits and video references cost $0.094 a second at 720p on the source plus the output.
Can Kling 4.0 render text?
On our title card, yes: it spelled 'MODA SUMMER SESSIONS' and 'Friday 12 June' correctly, where both Kling 3.0 tiers garbled the title. That is one run of one prompt.
What is Kling 4.0 Flash?
A faster version for frequent drafts, at 720p and up to 20 seconds. fal lists it in early access, not on the public API, as of 21 October 2026; we have not tested it.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
