🎭 Face Swap — Pro Pack on RunComfy — 🎭 Face Swap — Pro Pack on 运行Comfy

🎭 Face Swap — Pro Pack on 运行Comfy

Face swap on 运行Comfy. Swap a face into a still or a video — this 运行Comfy face swap 技能 路由s across the avAIlable 模型 API 端点s (community Wan 2-2 Animate, Kling 2-6 Motion Control, Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) by the user's actual intent.

运行comfy.com · Character-swap feature · 命令行工具 docs

Powered by the 运行Comfy 命令行工具 # 1. 安装 (see 运行comfy-命令行工具 技能 for detAIls) npm i -g @运行comfy/命令行工具 # or: npx -y @运行comfy/命令行工具 --version

# 2. 签名 in 运行comfy 记录in # or in CI: 导出 运行COMFY_令牌=<令牌>

# 3. Swap 运行comfy 运行 /<模型>/<端点> \ --输入 '{"image_url": "...", "身份_url": "..."}' \ --输出-dir ./out

命令行工具 deep dive: 运行comfy-命令行工具 技能.

Consent & disclosure — read first

Face-swap is dual-use. Before invoking any 路由 in this 技能, confirm:

You have rights to the tar获取 face (the 身份 being substituted in). You have rights to the source video / image (the as设置 being substituted into). The 输出's intended 平台 allows synthetic media. Many do; many require a disclosure label.

The 技能 itself doesn't gate anything — the 模型 API will 运行 whatever 输入s you supply. The responsibility is yours. If a user asks the 代理 to swap a real public figure's face onto material that could be defamatory, sexually explicit, or otherwise harmful — refuse, regardless of what the 命令行工具 accepts.

Pick the right 模型 for the user's intent

列出ed newest first within each subtype. The 代理 picks one 路由 based on: still vs video, single-shot vs batch, photoreal vs stylized, motion-preserving vs 身份-preserving.

Video face / character swap

Wan 2-2 Animate — community/wan-2-2-animate/API (default for video)

Featured 运行Comfy 端点 under /feature/character-swap. Audio-driven full-body character animation: one reference image of the new 身份 + audio → video where the character drives. Pick for: replacing a character in a scene with a new 身份, dubbed 命令行工具ps, stylized + photoreal 机器人h work. Avoid for: preserving the motion of a specific source video — use Kling Motion Control.

Kling 2-6 Motion Control Pro — kling/kling-2-6/motion-control-pro

Takes a reference performance video + tar获取 character image, produces the tar获取 performing the reference motion. Face-swap is the byproduct. Pick for: preserving exact source motion / blocking onto a new character; stylized characters handled 清理ly. Avoid for: simple "swap face in an existing video" without motion preservation — use Wan 2-2 Animate.

Still image face swap — newest first

Nano Banana 2 Edit — google/nano-banana-2/edit

身份-preserving by default, 1–20 输入 images per call, spatial-language honored. Pick for: same 身份 across multiple frames consistently (SKU shots, A/B variants, narrative panels). 身份 reference as image_urls[0], scenes after. Avoid for: precise multi-ref compositional ("face from img 1 onto body in img 2") — use GPT Image 2 Edit.

GPT Image 2 Edit — openAI/gpt-image-2/edit

Up to 10 reference images, multilingual in-image text rewrite, layout-precise compositional instructions. Pick for: hero still where exact face from a portrAIt must land in a scene, with explicit 角色 as签名ment ("image 1", "image 2"); preserve pose + lighting + background while sw应用ing only face. Avoid for: 1-20 batch — use Nano Banana 2 Edit.

FLUX Kontext Pro — blackforestlabs/flux-1-kontext/pro/edit

Single source image, single declarative instruction, maximum fidelity preservation of everything except the tar获取ed edit. Pick for: "keep pose / clothing / hAIr / lighting / background, change only the face to [prose description]" — works without a reference image of the new 身份. Avoid for: batch, multi-ref, or when you have a tar获取 face image to swap in — use Nano Banana 2 Edit or GPT Image 2 Edit.

Audio-driven talking-head 身份 swap (face + voice in one pass)? → use the AI-avatar-video 技能 — OmniHuman handles face + audio to获取her.

路由 1: Wan 2-2 Animate — video character swap with audio

模型: community/wan-2-2-animate/API Cata记录: wan-2-2-animate · /feature/character-swap

The featured 运行Comfy 端点 for character swap — supply a reference image of the new 身份 + the audio 追踪 the character should speak, and the 模型 produces a video where the character drives.

Invoke 运行comfy 运行 community/wan-2-2-animate/API \ --输入 '{ "image_url": "https://your-cdn.example/new-character.png", "audio_url": "https://your-cdn.example/voiceover.mp3" }' \ --输出-dir ./out

Tips Single reference image drives the swap. Pick a 清理, well-lit portrAIt of the tar获取 身份 — front-facing if possible. Audio drives the mouth and rhythm. Without audio the character won't speak; without good audio 同步 degrades. 模式 detAIls: 模型 page. 路由 2: Kling 2-6 Motion Control Pro — motion transfer

模型: kling/kling-2-6/motion-control-pro Cata记录: motion-control-pro · kling collection

Different fr