higgsfield-generate

Generate images, videos, 3D assets, and audio via the Higgsfield AI CLI.

Updated Jul 10, 2026
One-click install
npx skills add https://github.com/Kaleb-Rupe/aurora --skill higgsfield-generate-kaleb-rupe
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: higgsfield-generate
Source: https://github.com/Kaleb-Rupe/aurora/tree/main/growth/.agents/skills/higgsfield-generate
Command: npx skills add https://github.com/Kaleb-Rupe/aurora --skill higgsfield-generate-kaleb-rupe

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Choosing the right generative AI model and correctly submitting jobs for image, video, 3D, audio, or ad creation is complex and error-prone. This Skill wraps the Higgsfield CLI with model-routing guidance, media input handling, and Marketing Studio workflows so generation requests succeed on the first attempt. ## Core Features & Use Cases - Multi-modal generation: Create images (GPT Image 2, Nano Banana 2, Soul 2.0), videos (Seedance 2.0, Kling 3.0), 3D GLB assets (multi_image_to_3d), and audio (Seed Audio 1.0) through one CLI. - Marketing Studio ads: Produce branded UGC, unboxing, TV spot, and product showcase videos using avatars, products, hooks, settings, brand kits, and DTC ad formats. - Virality Predictor: Score a finished video's hook strength, attention, and retention via the brain_activity analysis model. - Use Case: A marketer provides a product URL and asks for a 15-second vertical UGC ad; the Skill fetches the product, picks an avatar, submits a marketing_studio_video job with --wait, and returns the result URL. ## Quick Start Ask the assistant to generate an image or video with a prompt, for example: generate a 12-second cinematic video of a neon city at dusk using Higgsfield.

Frequently Asked Questions about higgsfield-generate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a video with Higgsfield AI?

Run higgsfield generate create with a video model such as seedance_2_0, passing a prompt, optional start image, duration, and resolution, plus --wait to block until completion. The CLI prints the result URL when the job finishes.

Which Higgsfield model should I use for image generation?

GPT Image 2 is the default for high-fidelity general images, design, and on-image text. Use Nano Banana 2 for character or cartoon work, Recraft V4.1 for logos and vector-style graphics, and Soul 2.0 for aesthetic UGC or fashion editorial images.

Can I create a UGC ad video from a product URL?

Yes. Fetch the product with higgsfield marketing-studio products fetch --url, then run higgsfield generate create marketing_studio_video with the same URL, a mode like ugc, duration, and aspect ratio. The backend reuses the fetched product entity.

Does Higgsfield support image-to-3D model generation?

Yes, the multi_image_to_3d model creates an actual 3D mesh or GLB asset from 1 to 4 reference images passed with repeated --image flags. Add --should_texture true when the asset needs textures.

Why does my Higgsfield job fail with Session expired?

The Session expired error means your CLI credentials are no longer valid. Run higgsfield auth login interactively to re-authenticate, then retry the generation command.

When should I not use this generation skill?

Do not use it for Soul Character training (use higgsfield-soul-id), branded product photoshoots (use higgsfield-product-photoshoot), marketplace listing cards, or text/chat/TTS tasks, which are outside its scope.