As of: 2026-10-25 23:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/m/kling-4-0/

# Kling 4.0 (and 4.0 Flash)

Kuaishou, Kling 4 family. Status: current. Type: video generation. Released: 2026-10-15.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| fal.ai | `kling/v4/image-to-video` | https://queue.fal.run/kling/v4/image-to-video | https://fal.ai/models/kling/v4/image-to-video/api |
| fal.ai | `kling/v4/text-to-video` | https://queue.fal.run/kling/v4/text-to-video | https://fal.ai/models/kling/v4/text-to-video/api |
| fal.ai | `kling/v4/reference-to-video` |  | https://fal.ai/models/kling/v4/reference-to-video/api |
| fal.ai (Omni video edit) | `kling/v4/video-to-video/edit` |  | https://fal.ai/models/kling/v4/video-to-video/edit |
| fal.ai (4.0 Flash) | `kling/v4/flash/text-to-video` |  | https://fal.ai/models/kling/v4/flash/text-to-video |
| Web app |  |  | https://kling.ai |

## Pricing

- Per second 720p: $0.157
- Per second 1080p: $0.235
- Unit as published: USD per second of output on fal (kling/v4/text-to-video, image-to-video and reference-to-video, read 2026-10-25); reference-to-video with video references $0.094 / $0.141 per s, billed on input plus output length (minimum 2x the output). Kling's own API price not found
- Source: https://fal.ai/models/kling/v4/image-to-video
- Last checked: 2026-10-25

## Specs

- Input: text, image, video
- Output: video, audio
- Context window: not published
- Max output: not published
- Open weights: no
- Licence: proprietary

## Where it stands

- Newer video generation models in the registry since its release: 1:
  - FLUX 3 Max (2026-10-19) https://postcutoff.com/m/flux-3-max/
- Before it from Kuaishou: Kling 3.0 (VIDEO 3.0 / 3.0 Omni) (2026-02) https://postcutoff.com/m/kling-3-0/
- Closed weights; access is hosted only.

## What stands out

- 30-second single-pass clips: Native generation of up to 30 s in one pass (Kling 3.0: 15 s); fal's endpoints take 3-30 s, 720p or 1080p, with per-shot multi-prompts. Source: https://bernama.com/en/press/news.php?id=2622471
- Omni Reference with keyframes: Up to 15 reference assets and up to 10 keyframe inputs per generation (Kuaishou release); start and end frame on image-to-video. Source: https://bernama.com/en/press/news.php?id=2622471

## Notes

Unveiled 2026-09-28 (Lite/Flash for annual subscribers), official launch 2026-10-15 per Kuaishou's Oct 19 release, open to all users on klingai.com from ~Oct 16 (Pandaily: with audio, MCP and CLI hooks). Kling's own developer API for 4.0 was reported as 'coming soon' by third-party trackers (creaa.ai, wireflow, Oct 16); not checked on Kling's JS-rendered docs. fal's v4 endpoints (OpenAPI schema read 2026-10-25) expose no audio parameter, so audio via fal is not confirmed. Not yet on the Artificial Analysis text-to-video or image-to-video boards (read 2026-10-25), so it has no independent rank; Kling 3.0 Pro is the boards' 1000 anchor. 4K and 10-bit HDR availability reported inconsistently.

Kuaishou's Kling 4.0 video model: up to 30 s per pass, up to 15 references and 10 keyframes. Through fal it is an async queue call (submit, then poll the request id).

```bash
curl -X POST https://queue.fal.run/kling/v4/image-to-video -H "Authorization: Key $FAL_KEY" \
  -H "Content-Type: application/json" \
  -d '{"start_image_url":"https://example.com/character.png","prompt":"she turns and walks toward the camera","duration":5,"resolution":"1080p"}'
```

Sources: https://bernama.com/en/press/news.php?id=2622471 · https://fal.ai/models/kling/v4/image-to-video · https://fal.ai/models/kling/v4/text-to-video · https://fal.ai/models/kling/v4/reference-to-video · https://technode.com/2026/10/16/kling-ai-4-0-suggests-a-shift-in-ai-video-competition-from-visual-quality-to-directorial-control/ · https://creaa.ai/blog/kling-4

Model page: https://postcutoff.com/m/kling-4-0/