# Kling v3.0 Text-to-Video

Kling v3.0 Text-to-Video is the v3.0 text-to-video model in the Kling series. It produces multiple distinct scene segments from a single prompt at the v3 quality tier, with native multilingual audio.

- **Model ID:** `klingai/kling-v3.0-t2v`
- **Type:** video
- **Providers:** klingai
- **Context window:** Not applicable
- **Maximum output tokens:** 0
- **Pricing:** See the provider details below
- **Canonical page:** https://vercel.com/ai-gateway/models/kling-v3.0-t2v

## Supported parameters

- **Parameters:** `aspectRatio`, `resolution`, `duration`, `fps`
- **Operations:** `text-to-video`
- **Input modalities:** `text`
- **Output modalities:** `video`
- **Aspect ratios:** `16:9`, `9:16`, `1:1`
- **Resolutions:** `720p`, `1080p`, `4k`
- **Durations:** `3s`, `5s`, `10s`, `15s`
- **Frame rates:** `30 fps`
- **Maximum outputs per request:** Not reported

## Example

```ts
import { experimental_generateVideo as generateVideo } from 'ai';

const result = await generateVideo({
  model: 'klingai/kling-v3.0-t2v',
  prompt: 'A serene mountain lake at sunrise.'
});
```

## About

Kling v3.0 Text-to-Video introduces multi-shot generation as its signature feature. A single prompt can describe a multi-scene narrative. The model produces up to five coherent shots in one generation pass, each with its own visual composition and action. Total video duration runs up to 15 seconds across these shots, edited together as a continuous sequence. This eliminates the manual workflow of generating and stitching individual clips for multi-scene narratives.

The v3 generation tier improves visual quality in several areas. More realistic physics simulation governs object interactions, environmental elements, and secondary motion. Temporal consistency across frames is stronger. Native audio generation (multilingual speech in English, Chinese, Japanese, Korean, Spanish, and others, plus action sound effects and ambient audio) integrates into the same inference call.

For narrative-driven content production, advertising, and creative storytelling, v3.0 t2v reduces the number of sequential generation calls needed for a multi-scene video. Directing multiple shots from a single descriptive prompt also makes it well suited to AI-assisted storyboarding and pre-visualization workflows.

## What to consider

Multi-shot generation in v3.0 bills total output duration as the sum of each shot duration. Plan cost per asset with that in mind.

## When to use

### Best For
- Multi-scene narrative video from a single prompt: story sequences, ads with setup and payoff, or explainer progressions
- V3.0 text-to-video work where visual detail, motion quality, and audio matter more than turbo speed
- Content requiring multilingual narration or layered audio alongside generated visuals
- Creative pre-visualization and storyboarding for film or advertisement production

### Consider Alternatives When
- A reference image must anchor the subject appearance or visual style. Use v3.0 i2v
- A reference performance video defines the desired motion. Use motion control
- Cost and speed matter more than maximum quality and multi-shot capability. Use v2.5 Turbo t2v

## Best for

- **Multi-scene narrative video:** Story sequences, ads with setup and payoff, or explainer progressions from a single prompt
- **Quality-first text-to-video:** Work where visual detail, motion quality, and audio matter more than turbo speed
- **Multilingual narration:** Content requiring multilingual voice or layered audio alongside generated visuals
- **Creative pre-visualization:** Storyboarding for film or advertisement production

## Consider alternatives

- **Image-anchored subject:** A reference image must anchor the subject appearance or visual style, so use v3.0 i2v
- **Motion transfer required:** A reference performance video defines the desired motion, so use motion control
- **Cost and speed priority:** Cost and speed matter more than maximum quality and multi-shot capability, so use v2.5 Turbo t2v

## Frequently asked questions

### What is multi-shot generation in Kling v3.0 Text-to-Video?

A single prompt produces up to five distinct video shots in one inference pass. The shots form a continuous edited sequence. This enables narrative multi-scene videos without manual clip stitching.

### What is the maximum output duration?

Up to 15 seconds total across all shots, extended from 10 seconds in earlier Kling versions.

### What languages does the native audio generation support in v3.0?

English, Chinese, Japanese, Korean, and Spanish, including accented variants.

### How does v3.0 t2v differ from v2.6 t2v?

V3.0 adds multi-shot narrative generation, extends maximum duration to 15 seconds, improves physics simulation, and expands audio language support compared to v2.6.

### How is total video cost calculated for multi-shot output?

This page lists the current rates. Multiple providers can serve Kling v3.0 Text-to-Video, so AI Gateway surfaces live pricing rather than a single fixed figure.

### Does v3.0 t2v support vertical video output?

Yes. The 9:16 aspect ratio is supported alongside 16:9 and 1:1.

## Links

- [Model page](https://vercel.com/ai-gateway/models/kling-v3.0-t2v)
- [AI Gateway documentation](https://vercel.com/docs/ai-gateway)
- [Provider model documentation](https://app.klingai.com/global/release-notes/whbvu8hsip?type=dialog)
- [Provider pricing](https://klingai.com/global/dev/pricing?scrollTo=video)


If you encounter incorrect information, can’t find what you need, need a capability Vercel doesn’t currently offer, or have any other feedback to report, please do it by following the instructions at https://vercel.com/feedback.md.