KlingAI 2.5 Turbo Pro
Speed as a feature, motion as the product
- frameImages
- video
The "Turbo" in the name is the whole pitch. Kling 2.5 Turbo Pro is the fast lane of the Kling lineup on Runware - you trade some of the flagship's fidelity for markedly quicker renders and a lighter credit bill, and you get a node whose only required input is a prompt. For anyone whose bottleneck is "I need to see the motion before I commit to it," this is the node that turns video iteration into a loop instead of an event.
Why you'd reach for it
Because a slow loop is a broken workflow. The classic cloud-video frustration is paying flagship prices and waiting out long queues just to discover the motion was wrong. Turbo exists to make that cheap and quick: you explore the motion space, find the take that works, then (if you need the polish) re-render the keeper on a full-quality node. There's an argument that Turbo's output is good enough to be the final render for a lot of use cases anyway - social clips, storyboards, rough cuts.
How it works
Generated Runware node with the pack's shared base class: your positivePrompt (and optional frameImages, base64'd to PNG data URIs) rides to the Runware REST API over the runware SDK, and the finished clip is downloaded and wrapped as a native ComfyUI VIDEO object. Cost and NSFW flag print on the node title bar. The speed lives on Runware's side - your machine is just the remote control.
Inputs that matter
positivePrompt- required. Text-to-video by default; addframeImagesfor image-to-video.frameImages(IMAGE) - optional first frame.duration- 5 or 10 seconds. Frames = duration × fps, and faster rendering makes 10s feel reasonable.width/height- 1024×1024 default.CFGScale- 0.5 default;negativePromptignored below guidance ~1.numberResults(1–4) - with Turbo pricing, rolling a few variations is actually affordable.outputFormat(MP4/WEBM/MOV),outputQuality(95).
Output: one video socket (VIDEO type).
Install and the API key
cd ComfyUI/custom_nodes
git clone https://github.com/Runware/ComfyUI-Runware
pip install -r ComfyUI-Runware/requirements.txt
Restart ComfyUI (or Runware via ComfyUI Manager). Key from the Runware dashboard → Settings → Runware API key, RUNWARE_API_KEY, or runware auth login.
Gotchas
- No key → the pack's one-line "No Runware API key" error;
RUNWARE_DEBUG=1for full tracebacks. frameImagesis anIMAGEtensor from Load Image, not a path.- Turbo is a fidelity trade, not a free lunch. If the model's motion coherence at 720p-class resolution isn't holding up for your shot, that's the trade showing - promote the final to a higher tier.
- Closed weights, per-run pricing, cost on the title bar - same as every Kling node.
2.5 Turbo Pro is the node you'll actually use daily because it makes the feedback loop fast enough to be fun. Speed as a feature isn't marketing here; it's the difference between guessing and iterating.
Inputs (14)
| Name | Type | Default | Description |
|---|---|---|---|
| positivePrompt | STRING | Text prompt describing elements to include in the generated output. | |
| frameImagesopt | IMAGE | — | |
| negativePromptopt | STRING | Prompt to guide what to exclude from generation. Ignored when guidance is disabled (CFGScale ≤ 1). | |
| sizeopt | COMBO | 720p (1:1) | Output size. Only one of resolution or dimensions is sent. |
| CFGScaleopt | FLOAT | 0.500–1 | Guidance scale representing how closely the output will resemble the prompt. Higher values produce results more aligned with the prompt. |
| durationopt | COMBO | 5 | Length of the generated video in seconds. The total number of frames produced is determined by duration multiplied by the model's frame rate (fps). |
| numberResultsopt | INT | 11–4 | Number of results to generate. Each result uses a different seed, producing variations of the same parameters. |
| safetyopt | BOOLEAN | false | Enable to set safety. Off uses the model's default. |
| safety.checkContentopt | BOOLEAN | false | Enable or disable content safety checking. |
| safety.modeopt | COMBO | fast | Safety checking mode for video generation. |
| ttlopt | BOOLEAN | false | Enable to set ttl. Off uses the model's default. |
| ttl_valueopt | INT | 60 | Time-to-live (TTL) in seconds for generated content. Only applies when `outputType` is `URL`. |
| outputFormatopt | COMBO | MP4 | File format for the generated video. |
| outputQualityopt | INT | 9520–99 | Compression quality of the output. Higher values preserve quality but increase file size. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| video | VIDEO | — |