Type: Image
21 models, with aspect ratio, quality, resolution and images per prompt depending on the model.
Image and video batches in Higgsfield, with the model, its options and its reference files set for you.
Higgsfield has many models and each shows different options: some want a start frame and an end frame, others take several references, others none, and a few hide their settings behind a dropdown. Building a batch by hand means reconfiguring the page for every prompt.
Higgsfield AutoKuak knows the whole catalogue. When you choose a model the panel shows exactly its options and its reference slots, and on start it applies them on the page, checking one by one that they were accepted. Anything the site refuses because of your plan doesn’t break the batch: it’s logged and the run continues.
It’s the extension for producing with Higgsfield at volume: series of shots on the same model and settings, character variations, image batches for an edit. It uses your Higgsfield account and plan; the extension only handles the legwork.
The extension works on higgsfield.ai’s generation pages. If you’re outside one, the panel says so and unlocks by itself as soon as you enter.
“Paste manually”, with a blank line between prompts, or “From a tab”, cutting between two text markers in another open tab.
Image or Video, then the model. Once picked, the panel shows only the options that model really has; if it has none, it says so instead of faking a form.
In “Files + text” mode a block appears per slot, with the name the site itself uses, the types it accepts (photo, video or audio), how many, and whether it’s required.
Download quality, folder, prefix and automatic renaming, per group.
The extension applies the model and its settings, uploads the references, sends each prompt, waits for the result in the history and downloads it. The run view shows where each one stands.
Type and model are set per group; the input mode decides whether the prompt travels alone or with files.
21 models, with aspect ratio, quality, resolution and images per prompt depending on the model.
49 models, one clip per prompt, with duration, aspect, resolution and bitrate depending on the model.
The prompt travels alone. It’s the mode for models that take no reference files.
The prompt travels with each slot’s files. They upload once and serve the whole batch.
A different file per prompt, in order, into the model’s main slot.
Prompts are counted per tool over a rolling 24-hour window: each prompt stops counting 24 hours after it was sent.
| What | No account | Free account | Pro plan |
|---|---|---|---|
| Prompts to Higgsfield | 15 per 24 h | 30 per 24 h | Unlimited |
| Image and video, all 70 models | Yes | Yes | Yes |
| Reference files and automatic download | Yes | Yes | Yes |
Sign in with Google and start on the free plan. No card.