Kling 4.0 Text to Video: How to Prompt Shots Before Kling 5

Kling 5 isn't out yet. Learn how Kling 4.0 text to video works on Kling5 AI, which settings you can use today, and how to write a clear one-shot prompt.

A handwritten shot description dissolving into a row of glowing video frames on a dark desk, with a faint unused frame outline beside them
Table of contents

Kling 5 is not out yet as of November 17, 2026. If you want to run Kling text to video today, the newest model on Kling5 AI is Kling 4.0 text to video, and Kling 3.0 is also available. Kling5 AI is an independent site and is not affiliated with Kuaishou or Kling's app.

What does Kling 4.0 text to video do right now?

Kuaishou launched Kling 4.0 on October 15, 2026, according to Kuaishou's investor relations announcement dated October 17, 2026. In text-to-video mode, the model turns a written description into a single continuous video clip without requiring an image or video upload.

On Kling5 AI, Kling 4.0 text to video generates 3-to-30-second clips at 720p or 1080p across four aspect ratios. You can read our Kling 4.0 feature overview for a broader look at the release.

Kling5 AI's tool does not include every option listed in Kling's video capability map (updated October 20, 2026). Kling's app and API support 4K output, native audio, multi-shot storyboards, and Kling 4.0 Flash, while Kling5 AI keeps a simpler single-shot web workflow.

Setting or FeatureKling5 AI (Kling 4.0 Text to Video)Kling's app and API (Kling 4.0)
Duration3 to 30 seconds (1-second steps; default 5s)3 to 30 seconds
Resolution720p (default) or 1080p720P, 1080P, or 4K
Aspect ratios16:9 (default), 9:16, 1:1, 21:9Aspect controls supported per endpoint
Audio controlNo sound on/off switchNative audio supported
Shot structureSingle-prompt UI only (no storyboard or keyframe UI)Single-shot and multi-shot supported
Prompt limit2,500 charactersUnder 8,000 tokens (up to 26,000 English chars on API)
Kling 4.0 FlashNot availableAvailable (3 to 20 seconds, 720P only)

Several other features from Kling's app and API are also absent from Kling5 AI's Kling 4.0 page. There is no 4K toggle, no 10-bit HDR option, no two-minute Video Extend tool, no Elements character library, and no @Image1 or @image_1 tag syntax in text-to-video mode.

How do you generate a Kling 4.0 text to video clip on Kling5 AI?

Running a prompt on Kling5 AI takes five steps inside the Kling 4.0 tool. Before you start, make sure you are in text-to-video mode rather than image-to-video or reference-to-video on the same page.

  1. Write your prompt (required, up to 2,500 characters). Describe what happens in a single shot. Note that Kling's 4.0 Text-to-Video API documentation permits English-only prompts up to 26,000 characters and Kling's 4.0 Video User Manual cites a limit under 8,000 tokens, so do not paste an API-length script into Kling5 AI's 2,500-character box.
  2. Pick a duration from 3 to 30 seconds. You can step in 1-second increments, with 5 seconds selected by default.
  3. Select 720p or 1080p quality. The default is 720p, which is a practical starting point for testing pacing before switching to 1080p.
  4. Choose an aspect ratio. Options are 16:9 (default), 9:16, 1:1, and 21:9.
  5. Review the credit cost and generate. The tool displays the exact credit cost for your chosen duration and resolution before you submit.

Because the Kling 4.0 tool on Kling5 AI has no sound toggle, your generated video will not include native audio controls. If your project requires generated sound, check the Kling 3.0 section below.

How to write a clear one-shot text-to-video prompt

Good Kling 4.0 text to video prompting relies on concrete visual details rather than buzzwords. According to Kling's 4.0 Video User Manual and prompt guide, prompts work best when you avoid abstract phrases like "premium feel" or vague negatives and instead describe what the camera can actually see.

A reliable one-shot structure starts with a single grounding sentence covering the subject, setting, main action, and mood. Kling's Text-to-Video API documentation gives a simple baseline example: "A man in a white shirt walks along a beach at sunset, in a realistic cinematic style."

From that baseline, layer in camera movement, lighting, and timing if the shot changes over several seconds. Kling's user manual recommends ordering multi-beat motion by timestamp ranges such as 0-2 s, 2-5 s, and 5 s-end.

Two habits from Kling's docs do not apply to Kling5 AI's text-to-video box. First, leave out dialogue, music, and sound-effect instructions because Kling 4.0 on Kling5 AI has no sound switch. Second, never include @Image1 or @Video1 tags in text-to-video mode, as those tags only work in reference-based workflows.

You can copy and adapt these short one-shot prompt templates directly inside the 2,500-character limit:

  • 5-second product scene (16:9): A matte ceramic pour-over dripper rests on a dark walnut counter in soft morning window light, realistic commercial style. Thin steam rises steadily from the coffee as a slow close-up camera dollies slightly to the left.
  • 8-second vertical street shot (9:16): A cyclist in a yellow rain jacket pedals down a wet cobblestone alley at dusk, handheld documentary style. 0-3 s: medium tracking shot from behind as neon shop signs reflect in puddles. 3-8 s: the cyclist glances toward a steaming street stall and slows smoothly to a stop.
  • 10-second widescreen nature beat (21:9): A red fox stands on a snow-dusted pine ridge under pale winter sunlight, naturalistic cinema style. 0-4 s: wide static shot as a breeze lifts fine snow off nearby branches. 4-10 s: the fox trots across the ridge from left to right while the camera pans slowly to follow.
  • 6-second square studio shot (1:1): A potter shapes a wet clay bowl on a spinning wheel inside a sunlit workshop, crisp 35mm film look. Overhead close-up focused on clay-covered hands lifting the rim steadily without sudden cuts.

For a repeatable shot-plan formula and a wider set of prompt patterns, see our Kling 4.0 prompts guide.

When is text-to-video the wrong mode?

Text-to-video creates every pixel from words alone, just as Kling's Text-to-Video API endpoint handles text-only requests while reference assets go to a separate Omni endpoint. That makes text-to-video ideal for original scenes, mood pieces, and flexible compositions, but a poor fit when visual consistency is strict.

Switch out of text-to-video whenever you need an exact human face, a specific product label, or a locked starting and ending pose. On Kling5 AI's Kling 4.0 page, image-to-video mode requires a first frame and accepts an optional last frame, with output matching the aspect ratio of your uploaded frames.

Read our Kling 4.0 image-to-video guide for frame preparation tips. If you need a clean starting image to match your scene, you can generate one first with Kling5 AI's AI image generator.

Kling 4.0 on Kling5 AI also offers a reference-to-video mode for multi-asset consistency. That mode accepts up to 10 reference images and 5 reference videos, with at least one reference required.

Should you use Kling 3.0 text to video instead?

Kling 3.0 is still available on Kling5 AI and works well as a shorter, with-sound alternative when you do not need 20-to-30-second clips or ultra-wide framing. Check the credit cost displayed inside each tool before generating, as pricing adjusts by model and settings.

SettingKling 4.0 Text to Video (Kling5 AI)Kling 3.0 Text to Video (Kling5 AI)
Duration options3 to 30 seconds (1s steps; default 5s)3, 5, 10, or 15 seconds
Quality options720p (default) or 1080p720p (default) or 1080p
Sound toggleNoneOn or off (default off)
Aspect ratios16:9, 9:16, 1:1, 21:916:9, 9:16, 1:1 (no 21:9)
Prompt limitRequired, max 2,500 charsPrompt required

Kling 3.0 on Kling5 AI also includes image-to-video and motion-control modes—note that motion-control is available on Kling 3.0 only, not on Kling 4.0. While Kling's capability map lists 4K and multi-shot support for Kling 3.0 in the app and API, Kling5 AI caps Kling 3.0 text to video at 1080p and single-prompt clips.

Will Kling 5 change text to video?

As of November 17, 2026, Kuaishou has not announced or teased Kling 5 or Kling 5.0, and there are no specifications, release dates, or prices from Kuaishou. Any specs circulating on social media or third-party blogs are unverified and should not be treated as confirmed features.

Because Kuaishou has shared nothing publicly, there is no way to know whether a future release would extend clip length, change prompt limits, or add new camera controls. Kling5 AI plans to add it after an official release, with no timeline promised.

You can track updates on our release-date page or review our comparison of what Kling 4.0 does today.

Sources

Frequently asked questions

Is Kling 5 available for text to video yet?

No. As of November 17, 2026, Kuaishou has not announced or teased it, and no release date or specs exist. Today, the newest model you can run for Kling text to video on Kling5 AI is Kling 4.0, which launched on October 15, 2026. Kling5 AI expects to add the next model after an official launch.

How long can a Kling 4.0 text to video clip be on Kling5 AI?

On Kling5 AI, Kling 4.0 text to video supports durations from 3 to 30 seconds in 1-second increments, with 5 seconds set as the default. You can choose 720p or 1080p quality and 16:9, 9:16, 1:1, or 21:9 framing. Longer clips cost more credits, and the exact amount appears in the tool before you generate.

Does Kling 4.0 text to video on Kling5 AI include audio?

No. Although Kling's app and API list native audio support for Kling 4.0, the Kling 4.0 tool on Kling5 AI has no sound on or off switch. If you want generated sound directly inside a text-to-video clip on Kling5 AI, switch to Kling 3.0, which provides a sound toggle set to off by default.

Why is my prompt limited to 2,500 characters on Kling5 AI?

Kling5 AI enforces a 2,500-character maximum in its Kling 4.0 text-to-video prompt box to keep single-shot web inputs concise. Kling's documentation allows longer inputs-under 8,000 tokens in the user manual and up to 26,000 English characters on the API-so trim multi-beat prompts down to clear, visible actions before generating on Kling5 AI.

Can I use @Image1 tags or a character library in Kling 4.0 text to video?

No. On Kling5 AI, text-to-video mode accepts text only and does not support @Image1 or @image_1 tags, an Elements character library, or reference uploads. Those tag workflows belong to Kling's reference documentation. On Kling5 AI, switch to image-to-video for first and last frames or reference-to-video for reference images and clips.

Create with Kling 5 AI

Turn your idea into a video

Explore the video generator and start creating.

Generate a video
Kling 4.0text to videoKling 5Kling 3.0AI video

Written by

Kling 5 AI Team