Unifically LogoUnificAlly
Model logo

Higgsfield API

Image to Video

Cinematic videos with 50+ camera effects. Dolly shots, zooms, and drone perspectives.

Click or drag & dropPNG, JPG, WEBP, GIF · Max 100MB
Click or drag & dropPNG, JPG, WEBP, GIF · Max 100MB
Enhance Prompt
Whether to enhance/improve the prompt automatically
Output

Your generated video will appear here

Features

What Higgsfield API offers

Image-to-video from a required start_image_url plus descriptive prompt
Optional end_image_url for first and last frame control
Optional motion_id UUID to select a motion preset from Higgsfield's catalog
Three throughput tiers: Lite (slowest, baseline quality), Turbo (faster, mid quality), Standard (priority queue, highest quality per variant notes)
Optional enhance_prompt to let Higgsfield rewrite your wording before rendering
Seed support (0 through 999999) for reproducible rolls
Billing keys higgsfield-ai/lite, higgsfield-ai/turbo, and higgsfield-ai/standard depending on the variant

Use cases

Built for

Primary

Hero product spins - Animate still pack shots when you already have the key frame

#2

Real estate flythrough teasers - Motion templates instead of full 3D camera paths

#3

Music promos - Pair still artwork with kinetic motion without manual AE templates

#4

Social loops - Short motion pieces anchored on a single strong frame

#5

Story beats - Optional end frame for matched transitions between two still comps

#6

Match-referenced motion - Reuse motion_id values so recurring campaigns feel consistent

FAQ

About Higgsfield API

Every call needs a textual prompt and a start_image_url. You can optionally pass an end_image_url, motion_id preset UUID, enhance_prompt flag, or numeric seed.

They are throughput and priority tiers described in the playground: Lite is slowest, Turbo offers a quicker turnaround with mid-tier fidelity, Standard targets the best quality with priority queue placement.

It is a UUID that references a curated motion recipe inside Higgsfield's system. Populate it when you want template camera paths instead of improvising motion from text alone.

Usage maps to higgsfield-ai/lite, higgsfield-ai/turbo, or higgsfield-ai/standard, matching whichever variant you select before calling generate.