KlingAI 1.6 Standard

Kling 1.6 was the generation where people stopped apologizing for the results, and Standard is the cheaper way to get in on that. Same model family as the Pro node, same basic shape - positivePrompt required, optional frameImages for image-to-video - but a lower resolution ceiling and a friendlier credit bill. If you're doing high-volume motion tests or iterating on a look, this is the tier that makes the arithmetic work.

Why you'd reach for it

Two reasons, and they compound. First, 1.6's motion coherence - the subject-turns-around-and-stays-themself thing that made it famous - is worth having. Second, you don't need a GPU for any of it. Kling is Kuaishou's closed API model, so local reproduction was never on the table; this node just hands the render to Runware's cloud and gets a finished clip back. When you're burning through draft after draft, Standard's per-run price is the difference between iterating freely and rationing credits.

How it works

Standard pack machinery, same as its siblings: the class is generated from the catalog, your frameImages tensor (if wired) is base64'd to PNG data URIs, everything rides to the Runware REST API over the runware SDK, and the clip returns as a native VIDEO object. The node title bar prints each run's cost, so you can watch the savings happen in real time.

Inputs that matter

• positivePrompt - required. Text-to-video straight out of the box.
• frameImages (IMAGE) - optional; plug in a first frame for image-to-video.
• duration - 5 or 10 seconds.
• width / height - 1024×1024 default. At Standard's resolution ceiling, don't overshoot the pixel budget; it costs credits and buys little.
• CFGScale - 0.5 default; negativePrompt only earns its keep above guidance 1.
• numberResults (1–4), outputFormat (MP4/WEBM/MOV), outputQuality (95).

Output: one video socket (VIDEO type).

Install and the API key

The pack installs once and every Kling node comes along:

cd ComfyUI/custom_nodes
git clone https://github.com/Runware/ComfyUI-Runware
pip install -r ComfyUI-Runware/requirements.txt

Restart (or ComfyUI Manager → Runware). Key from the dashboard via Settings → Runware API key, RUNWARE_API_KEY, or runware auth login.

Gotchas

• frameImages is an IMAGE tensor, not a filename - Load Image first.
• Closed, paid, per-run: Kuaishou's weights by the request, cost shown on the node.
• The VIDEO output type needs current ComfyUI; on older builds it degrades to a saved file path.
• Standard is the value tier, full stop. The 2.x and video 3.0 nodes will out-render it if your budget doesn't care.

Honestly, for motion tests and draft pipelines, Standard is the one to leave in the graph. It's the 1.6 magic at a price you can iterate on.