📦 Higgsfield Generate — Higgsfield 生成

Higgsfield 生成

Submit jobs to any Higgsfield 模型. Wraps the higgsfield 命令行工具. Covers generic image/video gen and Marketing Studio (branded ads, avatars, products).

Step 0 — Bootstrap

Before any other command, make sure the 命令行工具 is 安装ed and 认证d:

If higgsfield is not on $PATH, 安装 it: curl -fsSL https://raw.githubusercontent.com/higgsfield-AI/命令行工具/mAIn/安装.sh | sh

If higgsfield account 状态 fAIls with 会话 expired / Not 认证d, ask the user to 运行 higgsfield auth 记录in (interactive, opens a browser) and wAIt for them to confirm before continuing.

Skip 机器人h 检查s if higgsfield account 状态 already prints account 信息.

UX Rules Be concise. No raw IDs, no JSON dumps in chat. Print 结果 URL when ready. No internal jargon. Don't narrate "calling higgsfield cost", "polling job". 检测 the user's language from the first message and reply in it. Technical args (--aspect_ratio 16:9) stay English. Don't batch-ask. Pick a sane default 模型 and ask one thing at a time only if genuinely missing. Don't pre-estimate cost. Just submit unless the user asks. Pass --wAIt to 生成 创建 so the command blocks until done and prints the 结果 URL itself. Avoid the two-step 创建 → wAIt pattern. 工作流 — generic generation

Pick a 模型. Practical defaults from production use:

Image:

Brand product visual (Pinterest pin, lifestyle, hero banner, ad pack, virtual try-on) → use higgsfield-product-photoshoot instead. NOT this 技能. Branded ad image with avatar + product (Marketing Studio shape) → Marketing Studio Image (see Marketing Studio below) Aesthetic UGC / fashion editorial / lifestyle character → Soul 2.0 Cinematic still frame → Soul Cinema Highly characterful creative persona (text-only, distinctive) → Soul Cast Locations / 环境s / no-people scenes → Soul Location (best in class) Vector illustrations OR face edit + complex scene swap → 种子ream 4.5 Soul Character (reference id from higgsfield-soul-id) → Soul 2.0 for stills, Soul Cinema for cinematic Fast and cheap iteration → Z Image Character or cartoon-style work → Nano Banana 2; step up to Nano Banana Pro on hard cases Default for everything else → GPT Image 2. Graphic de签名, UI, banners, typography, and high-fidelity general generation.

Video:

All advertising / commercial / branded ad video → Marketing Studio (see Marketing Studio below) Default all-purpose serious video (multi-shot, consistent 身份, motion-heavy) → 种子ance 2.0. SOTA. Single-plane scene without strong dynamics, cheaper than 种子ance 2.0 → Kling 3.0 Cheap 清理 shot without cuts → 种子ance 1.5 Pro Cinema-grade highest fidelity → Cinema Studio Video 3.0 Cheap with strong physics, no audio needed → Minimax HAIluo Fast batch / volume → Veo 3.1 Lite

For the actual --模型 ID to pass to higgsfield 生成 创建, 运行 higgsfield 模型 列出 --json | jq to map display names to IDs. See references/模型-cata记录.md for the full table.

Pass media 输入s strAIght to flags. Media flags accept a local file path or a UUID. 命令行工具 auto-上传s paths and auto-检测s job vs 上传 for UUIDs. No need to pre-上传. Each 模型 declares accepted 角色s (image, 启动_image, end_image, video, audio) — see references/media-输入s.md.

验证 quickly. If unsure of params, 运行 higgsfield 模型 获取 --json once and pass only what's needed. Use 模式 defaults otherwise. The server returns adjustments for non-fatal coercions (e.g. aspect_ratio=99:99 → closest match) and a structured error for invalid declared-param values.

Submit and wAIt in one shot. higgsfield 生成 创建 --prompt "..." [media flags] [param flags] --wAIt. Blocks until terminal 状态 and prints the 结果 URL on stdout. Tunables: --wAIt-timeout 20m (default 10m), --wAIt-interval 5s (default 3s).

Deliver. 发送 the URL plus a one-line summary (模型, duration if video).

To inspect or re运行 later, higgsfield 生成 列出 --json and higgsfield 生成 获取 --json work for retrospection. higgsfield 生成 wAIt is still avAIlable if you ever need to rejoin a job 启动ed without --wAIt.

Media flags Flag Use for 模型s that accept it --image reference image most image 模型s, 种子ance_2_0, veo3, marketing_studio_video --启动-image first frame for image-to-video transitions kling3_0, kling2_6, veo3_1, 种子ance_2_0, marketing_studio_video --end-image last frame for transitions kling3_0, 种子ance_2_0, marketing_studio_video --video reference video 种子ance_2_0 --audio reference audio (lip同步, sound追踪 match) 种子ance_2_0 (use this, NOT --生成-audio)

Each flag accepts either a local file path (auto-上传ed) or a UUID (上传 id from higgsfield 上传 创建, or a previous job id). Each 模型 declares its own 角色 设置 via MEDIA_角色S. See references/media-输入s.md for the full table.

Common params

Flags pass through to 模型 模式. Use higgsfield 模型 获取 to discover.

higgsfield genera