Kling 2.6 Text to Video with Audio
Describe a scene, get sound with it
- VIDEO
The Kling Text to Video with Audio node (Kling 2.6) is the text-only sibling of the 2.6 image-to-video-with-audio node, and it's the one that demonstrates Kling's audio trick most purely: type a scene, get back a clip with matching sound already in it. Rain, footsteps, a door creaking - the model generates the audio to go with what it renders. That's a real capability gap versus most video models, and it's why this node (and the whole "with audio" line) exists.
The pipeline is the standard partner-node shape. Your prompt goes to Kling's servers through your Comfy account, generation runs in the cloud, and the finished VIDEO returns when the task completes. No local model, no downloads - billed per generation to Comfy account credits, with audio roughly doubling the cost of a silent clip, and the node's price badge reflects that live.
The inputs that matter:
- prompt - the positive text prompt. Up to 2,500 characters, and since you're generating audio too, describing the soundscape helps: "ocean waves crashing, seagulls in the distance" gets you a more deliberate audio bed than "beach."
- model_name - fixed to
kling-v2-6. - mode -
prois the only option. - aspect_ratio - 16:9, 9:16, or 1:1.
- duration - 5 or 10 seconds.
- generate_audio - the toggle that matters, on by default. Flip it off for silent footage at a lower price.
Single VIDEO output into Save Video or Preview Video.
The practical knobs are few, so the decisions are: how long, what ratio, and whether to pay for sound. The gotcha is cost discipline - with the audio toggle defaulting to on, a batch of re-rolls quietly bills as if every attempt were premium. If you're iterating on motion first, kill generate_audio until you're happy, then turn it on for the keeper. Also keep the model line in perspective: 2.6 sits below Kling 3.0, which generates native audio on its turbo model and can't even turn it off. If you're choosing between the two, the 3.0 path (KlingVideoNode) is where the newer audio-video work is happening.
Standard partner-node housekeeping: be signed in to your Comfy account or the node fails immediately, and if the API-node family litters your search, --disable-api-nodes removes them all.
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| model_name | COMBO | 1 options: kling-v2-6 | |
| prompt | STRING | Positive text prompt. | |
| mode | COMBO | 1 options: pro | |
| aspect_ratio | COMBO | 3 options: 16:9, 9:16, 1:1 | |
| duration | COMBO | 2 options: 5, 10 | |
| generate_audio | BOOLEAN | true | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| VIDEO | VIDEO | — |