Kling 3.0 API: text or image to video, limits and price

Yes, Kling 3.0 has an API. Kling sells it through its own developer API, and multi-model APIs serve it too: on Sume it is the catalog id kling-3 ("Kling Video v3 Pro") on POST /v1/videos, which makes 4–15 second clips from a prompt or from a first and last frame, at 16:9, 9:16 or 1:1, with an optional audio track billed at a higher per-second rate.

Sume facts come from the Video generation docs, the Sume API reference, and the model catalog and pricing code behind GET /v1/videos/models, read on 2026-09-28; "current code" marks behavior read from the API code. Kling's own facts come from its API docs, read the same day.

How do I call Kling 3.0 from code?

On Sume, send the bare id kling-3 as model. Model ids carry no provider prefix, and switching to another catalog model is a change to that one field.

• Submit the request. The answer is 202 Accepted with a job id, a polling_url, and status: pending.
• Poll GET /v1/videos/{id} until status is completed (or failed / cancelled). The docs suggest about 30 seconds between polls; generation takes 30 seconds to several minutes. A callback_url (HTTPS) replaces polling with a webhook.
• Download from GET /v1/videos/{id}/content with the same API key. It redirects to the file, so curl needs -L.
• Send an Idempotency-Key header. A retry with the same key and the same body returns the original job instead of starting a second one.

curl -X POST "https://api.sume.com/v1/videos" \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kling-orbit-001" \
  -d '{
    "model": "kling-3",
    "prompt": "Cinematic product orbit on a clean table.",
    "duration": 5,
    "aspect_ratio": "16:9",
    "generate_audio": true
  }'

What does kling-3 accept on Sume?

These are the catalog values for kling-3. Confirm them with GET /v1/videos/models before you build on them.

How do I use Kling 3.0 for image to video?

Add frame_images to the same request: one entry with frame_type: "first_frame", and a last_frame entry if the clip must end on a second picture. A last frame without a first frame is refused. Images must be public HTTPS URLs. In current code, aspect_ratio is passed to the model only for text-to-video, so crop the first frame to the shape you want. Image to video with first and last frames covers frames across models.

How much does the Kling 3.0 API cost on Sume?

Sume bills video at the provider's list price × 1.25, reserved when you submit. For kling-3 that is $0.14 per second silent and $0.21 per second with audio, so a 10-second clip reserves $1.40 or $2.10, plus a 5.5% agent fee by default. Audio is on unless you send generate_audio: false, so leaving the field out bills the audio rate. pricing_skus lists both rates as per-video-second and per-video-second-audio.

The rate has no resolution tier, and in current code resolution is checked against the catalog but not passed on to the model, so don't count on that field to set Kling's output size on Sume.

Can I call Kling 3.0 directly from Kling?

Yes. Kling's own API takes a Bearer API key from the Kling AI console and prices Kling 3.0 per second in Units; Sume vs Kling compares its plans, prices and apps with Sume. The request itself looks like this.