Aurora v1 Fast
The same audio-to-video, on a budget clock
- image
- video
Aurora v1 Fast is the cheaper, quicker sibling of Aurora v1. Same audio-driven image-to-video trick - still in, soundtrack in, MP4 out - but the "Fast" in the name is the whole point: Runware routes it to a leaner checkpoint (creatify:aurora@fast) that trades a little final polish for speed and a smaller bill. If you're iterating on timing, lip-sync takes, or shot variations, this is the node to burn through your first ten attempts on.
The honest framing: Fast is for iteration, the full model is for the deliverable. Anyone who's done paid video work knows that rhythm - cheap passes until the take is right, then one expensive clean render. This node is the cheap pass, and it shares every socket with its full sibling, so you can swap them in and out of the same workflow without rewiring anything.
What you set
image(required) - the starting frame as anIMAGE. Same input as the non-Fast node.audio(required) - the track, as a UUID or URL. Hosted URL is the low-friction path.positivePrompt- optional guidance for what the motion shows. This is an audio-driven model; the prompt is seasoning, not the recipe.CFGScale- defaults to 1. It's low-guidance by design.safety.mode-none/fast/fullcontent check, off unless you need it.outputFormat/outputQuality- MP4 / WEBM / MOV, and 20–99 compression.
Output is a video (VIDEO) that drops straight into Preview Video or Save Video. Up to 4 results per run via numberResults.
How it works
Mechanically it's identical to every Runware node: build a request with the model AIR and taskType: videoInference, send it over REST through the SDK, pull the rendered file back as a native VIDEO. The only difference from Aurora v1 is which checkpoint Runware spins up on the other end. The title bar shows the per-run cost, which is the real way to feel the "Fast" difference - Fast runs should print meaningfully lower numbers.
Installing
ComfyUI Manager → search Runware → install → restart. Manual:
cd ComfyUI/custom_nodes
git clone https://github.com/Runware/ComfyUI-Runware
pip install -r ComfyUI-Runware/requirements.txt
No model downloads; the pack needs runware-sdk, pillow, and soundfile. API key via Settings → Runware API key, RUNWARE_API_KEY, or runware auth login.
Troubleshooting
Key-not-set errors dominate, same as the rest of the pack. If Fast's output is noticeably rougher than the full model, that's the trade you signed up for - it's the same input contract, so the fix is to render the final take on Aurora v1 proper. Audio not playing through? Check the URL/UUID is live, and that you didn't raise CFGScale and fight the model's natural behavior.
Inputs (12)
| Name | Type | Default | Description |
|---|---|---|---|
| image | IMAGE | — | |
| audio | STRING | Audio input (UUID or URL). | |
| positivePromptopt | STRING | Text prompt describing elements to include in the generated output. | |
| CFGScaleopt | FLOAT | 1.000–10 | Guidance scale representing how closely the output will resemble the prompt. Higher values produce results more aligned with the prompt. |
| 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 | — |