WaveSpeedAI Kling 4.0 Video Generation

Generate and edit videos with Kuaishou's Kling 4.0 on the WaveSpeed AI platform. Text, image and reference inputs produce 3-30 second clips at 720p or 1080p with synchronized stereo audio; video edit rewrites an existing clip and video extend appends a new segment.

Setup

Install the open-source CLI once and sign in; the CLI stores the key, so never ask the user to paste an API key into the chat:

npm install -g @wavespeed/cli
wavespeed login          # opens https://wavespeed.ai/accesskey and stores the key
wavespeed status         # confirms you are signed in

For CI or one-off shells, WAVESPEED_API_KEY in the environment also works.

Prefer MCP tools over shell commands? The same platform is exposed as an MCP server: hosted at https://api.wavespeed.ai/mcp (browser sign-in) or locally with npx -y @wavespeed/mcp@latest; every example below maps onto run_model with the same model id and input fields.

Quick Start

The shortcut commands use the same model and pick the endpoint from the inputs: wavespeed generate video "<prompt>" -m kling-4.0, and likewise wavespeed edit video and wavespeed extend video; the MCP tools generate_video, edit_video and extend_video take the same model value (confirm the keys with wavespeed generate video --help). Add --estimate --json first to see the chosen endpoint, final input and price without submitting.

wavespeed generate video "A paper boat drifting down a rain-soaked street" -m kling-4.0 --estimate --json

Text-to-Video

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/text-to-video \
  -p "A golden retriever running through a field of sunflowers at sunset, slow tracking shot" \
  --json | jq -r '.outputs[0]')

Image-to-Video

The image parameter accepts an image URL. If you have a local file, pass it with the @path marker and the CLI uploads it and substitutes the hosted URL.

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/image-to-video \
  -i image=@./photo.png \
  -p "The person slowly turns and smiles at the camera" \
  --json | jq -r '.outputs[0]')

Existing URLs work as-is:

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/image-to-video \
  -i image="https://example.com/photo.jpg" \
  -p "The person slowly turns and smiles at the camera" \
  --json | jq -r '.outputs[0]')

API Endpoints

duration is an integer from 3 to 30 seconds on every endpoint, and resolution is 720p or 1080p.

Text-to-Video

Model ID: kwaivgi/kling-v4.0/text-to-video

Parameters

Example

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/text-to-video \
  -p "A timelapse of a city skyline transitioning from day to night, cinematic slow pan" \
  -i aspect_ratio="21:9" \
  -i duration=10 \
  -i resolution="1080p" \
  --json | jq -r '.outputs[0]')

Image-to-Video

Model ID: kwaivgi/kling-v4.0/image-to-video

Animates a first frame, optionally toward a last frame and through ordered in-between keyframes.

Parameters

Example

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/image-to-video \
  -i image=@./sketch-1.png \
  -i keyframes='["@./sketch-2.png", "@./sketch-3.png"]' \
  -i last_image=@./sketch-4.png \
  -p "A seed sprouts, grows into a sapling and blossoms, steady camera" \
  -i duration=10 \
  --json | jq -r '.outputs[0]')

Reference-to-Video

Model ID: kwaivgi/kling-v4.0/reference-to-video

Builds a new clip that follows the subjects, motion and camera you point to. In the prompt, refer to inputs as @Image1, @Image2, ..., @Video1, ... and @Audio1, ... in array order.

Parameters

Example

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/reference-to-video \
  -i images='["@./character.png", "@./product.png"]' \
  -p "@Image1 picks up @Image2 from a cafe table and shows it to the camera, warm morning light" \
  -i aspect_ratio="9:16" \
  --json | jq -r '.outputs[0]')

Video Edit

Model ID: kwaivgi/kling-v4.0/video-edit

Keeps the scene, subject and motion of a 3-30 second video and changes what you describe (longer videos are trimmed to 30 s). Refer to the video as @Video1, to images as @Image1, ... and audios as @Audio1, ...

Parameters

Example

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/video-edit \
  -i video=@./street.mp4 \
  -i images='["@./red-jacket.png"]' \
  -p "Dress the man in @Video1 in the jacket from @Image1, keep everything else unchanged" \
  --json | jq -r '.outputs[0]')

Video Extend

Model ID: kwaivgi/kling-v4.0/video-extend

Generates a new segment from the input video's last frame and returns the original and the new segment joined; audio is generated with the picture.

Parameters

Example

OUTPUT_URL=$(wavespeed run kwaivgi/kling-v4.0/video-extend \
  -i video=@./clip.mp4 \
  -p "The camera pulls back to reveal the whole valley as the sun sets" \
  -i duration=8 \
  --json | jq -r '.outputs[0]')

Pricing

Estimates for a 5-second clip; cost scales with duration (10 s at 720p is $1.57). Video edit and reference-to-video with a reference video are also priced on the input video, so those rows assume a 5-second input; quote exact inputs with wavespeed price <model-id> ... --upload. Prices may change; see https://wavespeed.ai/models/<model-id> (e.g. text-to-video).

Prompt Tips

• Describe subject, action, camera movement and the sound you want; audio is generated with the picture
• Use keyframes and last_image in image-to-video to direct a sequence of story beats in order
• Name every reference by its tag (@Image1, @Video1, @Audio1) and say what each contributes
• For video edit, state what must stay unchanged; set keep_original_sound: false only when the new scene needs new audio
• Use 21:9 for ultrawide cinematic shots and 9:16 for vertical social video

CLI tips

# Inspect the live input schema before running (fields, enums, defaults)
wavespeed run kwaivgi/kling-v4.0/text-to-video -h

# Estimate the cost (no generation, no charge)
wavespeed price kwaivgi/kling-v4.0/text-to-video -p "..." -i key=value

# Save outputs to disk instead of only printing URLs
wavespeed run kwaivgi/kling-v4.0/text-to-video -p "..." --json --download "./out/{index}.{ext}"

# Local files: prefix the path with @ and the CLI uploads it and passes the hosted URL
wavespeed run kwaivgi/kling-v4.0/text-to-video -i <field>=@./local-file.png --json

# Recover a result if the run was interrupted (the id is in the --json output)
wavespeed show <id>

run --json prints { id, model, prompt, outputs: [url, ...], saved: [path, ...], elapsed_ms, raw }. Read outputs[0] for the result URL.

A run that fails with a gateway or network error (502/503/504, connection reset, timeout) does NOT mean the task failed, and the submit request is never retried automatically, so re-running can create a duplicate paid task. Do not re-run immediately. If the error contains a task id (task_id / prediction id), keep checking it with wavespeed show <id> --json until its status is terminal (completed, failed, cancelled, timeout, deleted). If no id was returned, check recent predictions with wavespeed history --json (add --model <id>; default window is the last 24h) and re-run once only if none matches. A client-side wait timeout or an interrupted run does not mean the task failed: tasks can run for a long time, so do not give up after a fixed time.

Security constraints

• Never ask for the key in chat: wavespeed login handles auth; if wavespeed status says signed out, ask the user to run it.
• Local files only via @: bare paths are passed through untouched and the model will reject them. Only @-prefixed values upload.
• No arbitrary URL loading: only pass media URLs the user provided or that came back from a previous run.
• Input validation: only pass parameters documented above; confirm with wavespeed run <model> -h when unsure.