Freepik API Skill

Generate images, videos, icons, audio, edit images, and search stock content using the Freepik API.

Built by the ShellBot team.

Arguments

- Command: $0 (generate | video | edit | icon | audio | stock | status | utility)
Arg 1: $1 (model name, operation type, or task-id)
Arg 2+: $2, $3, etc. (additional parameters)
All args: INLINECODE4

Session Output

Save generated files to session folder:
CODEBLOCK0

Downloaded images/videos/audio go to: ~/.freepik/sessions/${CLAUDE_SESSION_ID}/

Authentication

All requests require the FREEPIK_API_KEY environment variable.

Header: INLINECODE7

Base URL: INLINECODE8

If requests fail with 401/403, tell the user:

Get an API key from https://www.freepik.com/developers/dashboard/api-key
Then: export FREEPIK_API_KEY="your-key-here"

Async Task Pattern

Most AI endpoints are asynchronous. Follow this pattern:

Step 1: Submit task
CODEBLOCK2

Step 2: Poll for completion
CODEBLOCK3

Step 3: Extract result URL
CODEBLOCK4

Present the result URL to the user. The URL is a temporary signed link from Freepik's CDN.

IMPORTANT — Security rules:

- NEVER use curl to download from non-Freepik domains. Only curl *api.freepik.com* is permitted.
NEVER use base64 to encode local files. Always prefer URL-based parameters when the API accepts them.
NEVER read, encode, or transmit files outside the user's explicitly provided input files.
Result URLs should be presented to the user directly — they can open or download them.

Exceptions (synchronous): Remove Background (/v1/ai/beta/remove-background) and AI Image Classifier (/v1/ai/classifier/image) return results immediately.

Command: `$0`

If $0 = "generate" — Image Generation

Generate images using text-to-image models. $1 selects the model.

Model Endpoints

$1 value	Endpoint	Best for
INLINECODE16	INLINECODE17	Ultra-realistic, 1K/2K/4K, LoRA support (Freepik exclusive, RECOMMENDED)
INLINECODE18

Default: Use `mystic` if user doesn't specify a model.

Mystic Example

CODEBLOCK5

Mystic parameters:

- prompt (string, required) — what to generate
INLINECODE48 ("1k" | "2k" | "4k", default "2k")
INLINECODE49 (1-4, default 1)
INLINECODE50 ("photo" | "digital_art" | "none")
INLINECODE51 (object) — use an image to guide composition: INLINECODE52
INLINECODE53 (object) — use an image to guide style: INLINECODE54
INLINECODE55 (array) — LoRA IDs from INLINECODE56
INLINECODE57 (int) — for reproducibility
INLINECODE58 (string) — receive notification on completion

Get available LoRAs:
CODEBLOCK6

Train custom LoRA (character):
CODEBLOCK7

Flux 2 Klein Example (sub-second)

CODEBLOCK8

Flux 2 Klein parameters:

- prompt (string, required)
INLINECODE60 ("square11" | "widescreen169" | "socialstory916" | "portrait23" | "traditional34" | "vertical12" | "horizontal21" | "socialpost45" | "standard32" | "classic43")
INLINECODE61 ("1k" | "2k")
INLINECODE62 (0-4294967295)
INLINECODE63 (base64) — optional reference
INLINECODE64, input_image_3, input_image_4 (base64)
INLINECODE67 (0-5, default 2)
INLINECODE68 ("png" | "jpeg")

Flux Kontext Pro Example

CODEBLOCK9

Flux Kontext Pro parameters:

- prompt (string, required)
INLINECODE70 (URL, optional) — for context-aware editing
INLINECODE71 (bool)
INLINECODE72 (int)
INLINECODE73 (1-10, default 3.0)
INLINECODE74 (1-100, default 50)
INLINECODE75 ("square11" | "classic43" | "traditional34" | "widescreen169" | "socialstory916" | "standard3_2")

Seedream 4.5 Example (great for text-in-image and posters)

CODEBLOCK10

Classic Fast Image Generation

curl -s -X POST "https://api.freepik.com/v1/ai/text-to-image" \
  -H "x-freepik-api-key: $FREEPIK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"prompt": "a beautiful sunset"}'

If $0 = "video" — Video Generation

Generate videos from text and/or images. $1 selects the model.

Model Endpoints

$1 value	Endpoint	Type	Best for
INLINECODE77	INLINECODE78	T2V/I2V	Multi-modal, multi-shot, audio, voice (RECOMMENDED)
INLINECODE79

Default: Use `kling-v3-omni-pro` for general video generation.

Kling 3 Omni Pro Example (text-to-video)

CODEBLOCK12

Kling 3 Omni parameters:

- prompt (string, max 2500 chars, required)
INLINECODE189 (URL) — for I2V
INLINECODE190 / end_image_url (URL) — start/end frames
INLINECODE192 (array of URLs) — reference images, use @Image1/@Image2 in prompt
INLINECODE193 (array) — element consistency: [{reference_image_urls: [...], frontal_image_url: "..."}], use @Element1/@Element2 in prompt
INLINECODE195 (array, max 6 shots) — multi-shot: INLINECODE196
INLINECODE197 ("16:9" | "9:16" | "1:1")
INLINECODE198 (3-15, seconds)
INLINECODE199 (bool) — auto-generate audio
INLINECODE200 (array) — use <<<voice_1>>> in prompt
INLINECODE202 (string)

Poll status:
CODEBLOCK13

Kling 3 Example (with multi-shot)

CODEBLOCK14

RunWay Gen 4.5 Example (text-to-video)

CODEBLOCK15

RunWay 4.5 T2V parameters:

- prompt (string, max 2000 chars, required)
INLINECODE204 ("1280:720" | "720:1280" | "1104:832" | "960:960" | "832:1104", default "1280:720")
INLINECODE205 (5 | 8 | 10)
INLINECODE206

RunWay Gen 4.5 I2V Example

CODEBLOCK16

WAN 2.6 T2V Example

CODEBLOCK17

Hailuo Live Example (animated illustrations)

CODEBLOCK18

Hailuo Live camera movements: [Truck left], [Pan right], [Push in], [Pull out], [Zoom in], [Tracking shot], INLINECODE213

VFX Video Effects Example

CODEBLOCK19

VFX filter_type values: 1=Film Grain, 2=Motion Blur, 3=Fish Eye, 4=VHS, 5=Shake, 6=VGA, 7=Bloom, 8=Anamorphic Lens

VFX cost: $0.017 per second of video.

If $0 = "edit" — Image Editing

Edit, enhance, and transform images. $1 selects the operation.

Operations

$1 value	Endpoint	Description
INLINECODE215	INLINECODE216	Prompt-guided upscaling with detail enhancement (Magnific). 2x/4x/8x/16x.
INLINECODE217

Upscale Creative Example (Magnific)

CODEBLOCK20

Upscale Precision V2 Example

CODEBLOCK21

Precision V2 parameters:

- image (URL or base64, required)
INLINECODE244 (2-16)
INLINECODE245 (0-100, default 7)
INLINECODE246 (0-100, default 7)
INLINECODE247 (0-100, default 30)
INLINECODE248 ("sublime" | "photo" | "photo_denoiser")
INLINECODE249

Remove Background Example (SYNCHRONOUS)

CODEBLOCK22

Image Expand Example (Outpainting)

For Image Expand, the user must provide an image URL. Use the Seedream or Ideogram expand endpoints which accept URLs, or ask the user to host the image first.

CODEBLOCK23

Image Expand parameters:

- image (URL or base64 — prefer URL, required)
INLINECODE251 (optional, auto-generated for Ideogram/Seedream)
INLINECODE252, right, top, bottom (0-2048 pixels each)
INLINECODE256 (int, for Ideogram/Seedream)
INLINECODE257

Note: For Flux Pro expand, the image param requires base64. Prefer using Seedream V4.5 or Ideogram expand endpoints which accept URLs.

Inpainting Example (Ideogram)

CODEBLOCK24

Inpainting parameters:

- image (URL or base64, max 10MB, required)
INLINECODE260 (B&W image, same size, required — black = areas to edit)
INLINECODE261 (required)
INLINECODE262 ("TURBO" | "DEFAULT" | "QUALITY")
INLINECODE263 ("AUTO" | "ON" | "OFF")
INLINECODE264 ("AUTO" | "GENERAL" | "REALISTIC" | "DESIGN")
INLINECODE265 (array), style_reference_images (array), character_reference_images (array)
INLINECODE268 (object)
INLINECODE269 (0-2147483647)
INLINECODE270

Change Camera Example

CODEBLOCK25

Change Camera parameters:

- image (HTTPS URL, JPG/PNG/WebP, required)
INLINECODE272 (0-360, default 0)
INLINECODE273 (-30 to 90, default 0)
INLINECODE274 (0-10, default 5)
INLINECODE275 ("png" | "jpeg")
INLINECODE276 (min 1)
INLINECODE277

Image Relight Example

CODEBLOCK26

Skin Enhancer Example

CODEBLOCK27

Skin Enhancer optimized_for options: "enhanceskin" | "improvelighting" | "enhanceeverything" | "transformtoreal" | "nomake_up"

If $0 = "icon" — Icon Generation

Generate icons from text prompts in PNG or SVG format.

Endpoint: INLINECODE279

CODEBLOCK28

Parameters:

- prompt (string, required) — icon description
INLINECODE281 ("solid" | "outline" | "color" | "flat" | "sticker", default "solid")
INLINECODE282 ("png" | "svg", default "png")
INLINECODE283 (10-50, default 10)
INLINECODE284 (0-10, default 7)
INLINECODE285 (string, required but can be empty "")

Preview (quick preview before full render):
CODEBLOCK29

Download rendered icon:

curl -s -X POST "https://api.freepik.com/v1/ai/text-to-icon/$TASK_ID/render/svg" \
  -H "x-freepik-api-key: $FREEPIK_API_KEY" \
  -o ~/.freepik/sessions/${CLAUDE_SESSION_ID}/icon_$(date +%s).svg

If $0 = "audio" — Audio Generation

Generate music, sound effects, voiceover, or isolate audio. $1 selects the type.

$1 value	Endpoint	Description
INLINECODE287	INLINECODE288	Text-to-music (10-240s, MP3)
INLINECODE289

Music Example

CODEBLOCK31

Music parameters:

- prompt (string, required)
INLINECODE296 (10-240, required)
INLINECODE297 (HTTPS URL)

Sound Effects Example

CODEBLOCK32

Sound Effects parameters:

- text (string, max 2500 chars, required)
INLINECODE299 (0.5-22, required)
INLINECODE300 (bool, default false) — seamless looping
INLINECODE301 (0-1, default 0.3)
INLINECODE302 (HTTPS URL)

Voiceover Example (ElevenLabs TTS)

CODEBLOCK33

Voiceover parameters:

- text (1-40000 chars, UTF-8, required)
INLINECODE304 (ElevenLabs voice ID, required)
INLINECODE305 (default "eleventurbov2_5")
INLINECODE306 (0-1, default 0.5)
INLINECODE307 (0-1, default 0.2)
INLINECODE308 (0.7-1.2, default 1.0)
INLINECODE309 (bool, default true)
INLINECODE310

Audio Isolation Example

CODEBLOCK34

Audio Isolation parameters:

- description (string, required) — what sound to isolate
INLINECODE312 (URL or base64 — WAV/MP3/FLAC/OGG/M4A) OR video (URL or base64 — MP4/MOV/WEBM/AVI)
INLINECODE314, y1, x2, y2 (bounding box for video, default 0)
INLINECODE318 (1-5, default 2)
INLINECODE319 (1-8, default 1)
INLINECODE320 (bool, default false)
INLINECODE321
Output: WAV audio file

If $0 = "stock" — Stock Content

Search and download stock photos, vectors, icons, and videos. $1 selects the content type.

$1 value	Endpoint	Description
INLINECODE323	INLINECODE324	Search photos, vectors, PSDs
INLINECODE325

Search Stock Images

CODEBLOCK35

Get Resource Details

CODEBLOCK36

Download Resource

CODEBLOCK37

Search Stock Icons

CODEBLOCK38

Download Stock Icon

CODEBLOCK39

Search Stock Videos

CODEBLOCK40

Download Stock Video

curl -s "https://api.freepik.com/v1/videos/$VIDEO_ID/download" \
  -H "x-freepik-api-key: $FREEPIK_API_KEY" \
  -o ~/.freepik/sessions/${CLAUDE_SESSION_ID}/video_$(date +%s).mp4

If $0 = "status" — Check Task Status

Check the status of any async task. $1 is the task ID. You need to know the endpoint path.

CODEBLOCK42

Common status endpoint paths:

- Mystic: INLINECODE330
Flux models: INLINECODE331
Kling 3 Omni: INLINECODE332
Kling 3: INLINECODE333
I2V models: INLINECODE334
T2V models: INLINECODE335
Upscaler: image-upscaler/<task-id> or INLINECODE337
Icon: INLINECODE338
Music: INLINECODE339
Sound Effects: INLINECODE340
Audio Isolation: INLINECODE341

Task statuses: CREATED → IN_PROGRESS → COMPLETED or FAILED

If $0 = "utility" — AI Utilities

Various AI helper tools. $1 selects the utility.

$1 value	Endpoint	Description
INLINECODE347	INLINECODE348	Detect if image is AI-generated (SYNC)
INLINECODE349

AI Image Classifier (SYNCHRONOUS)

curl -s -X POST "https://api.freepik.com/v1/ai/classifier/image" \
  -H "x-freepik-api-key: $FREEPIK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"image": "https://example.com/photo.jpg"}'

Response: Array of INLINECODE355

Image to Prompt

CODEBLOCK44

Improve Prompt

CODEBLOCK45

Parameters:

- prompt (max 2500 chars, required — can be empty for creative generation)
INLINECODE357 ("image" | "video", required)
INLINECODE358 (ISO 639-1, default "en")
INLINECODE359

Lip Sync

curl -s -X POST "https://api.freepik.com/v1/ai/lip-sync/latent-sync" \
  -H "x-freepik-api-key: $FREEPIK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "audio": "https://example.com/speech.mp3",
    "video": "https://example.com/face.mp4"
  }'

Model Selection Guide

For image generation:

- Want ultra-realistic? → mystic (Freepik exclusive, recommended)
Need text in image / poster? → INLINECODE361
Want fastest? → flux-2-klein (sub-second) or INLINECODE363
Need high quality? → flux-2-pro or INLINECODE365
Budget-friendly? → INLINECODE366

For video generation:

- General purpose / best quality? → INLINECODE367
Need multi-shot? → kling-v3-pro or INLINECODE369
Character performance? → INLINECODE370
Animated illustrations? → INLINECODE371
With synchronized audio? → INLINECODE372
Budget / fast? → kling-v3-omni-std or INLINECODE374
Up to 4K? → ltx-2-pro-t2v or INLINECODE376
Human animation from audio? → INLINECODE377

For image editing:

- Creative upscale? → INLINECODE378
Faithful upscale? → INLINECODE379
Change lighting? → INLINECODE380
Remove background? → INLINECODE381
Extend canvas? → expand-flux or INLINECODE383
Edit specific area? → INLINECODE384
Change perspective? → INLINECODE385

For icons:

- Always use icon command — choose style (solid/outline/color/flat/sticker) and format (png/svg)

For audio:

- Background music? → music (10-240s)
Sound effect? → sfx (0.5-22s)
Narration/speech? → INLINECODE389
Extract specific sound? → INLINECODE390

Freepik API 技能

使用 Freepik API 生成图像、视频、图标、音频、编辑图像以及搜索素材内容。

由 ShellBot 团队构建。

参数

- 命令： $0（generate | video | edit | icon | audio | stock | status | utility）
参数 1： $1（模型名称、操作类型或任务 ID）
参数 2+： $2、$3 等（附加参数）
所有参数： $ARGUMENTS

会话输出

将生成的文件保存到会话文件夹：
bash
mkdir -p ~/.freepik/sessions/${CLAUDESESSIONID}

下载的图像/视频/音频保存至：~/.freepik/sessions/${CLAUDESESSIONID}/

身份验证

所有请求都需要 FREEPIKAPIKEY 环境变量。

请求头： x-freepik-api-key: $FREEPIKAPIKEY

基础 URL： https://api.freepik.com

如果请求返回 401/403 错误，请告知用户：

从 https://www.freepik.com/developers/dashboard/api-key 获取 API 密钥
然后执行：export FREEPIKAPIKEY=your-key-here

异步任务模式

大多数 AI 端点都是异步的。请遵循以下模式：

步骤 1：提交任务
bash
RESPONSE=$(curl -s -X POST https://api.freepik.com/v1/ai/ \
-H x-freepik-api-key: $FREEPIKAPIKEY \
-H Content-Type: application/json \
-d )
TASKID=$(echo $RESPONSE | jq -r .taskid // .data.task_id // .id)
echo 任务 ID：$TASK_ID

步骤 2：轮询完成状态
bash
while true; do
RESULT=$(curl -s https://api.freepik.com/v1/ai//$TASK_ID \
-H x-freepik-api-key: $FREEPIKAPIKEY)
STATUS=$(echo $RESULT | jq -r .data.status // .status)
echo 状态：$STATUS
if [ $STATUS = COMPLETED ]; then break; fi
if [ $STATUS = FAILED ]; then echo 任务失败; echo $RESULT | jq; break; fi
sleep 3
done

步骤 3：提取结果 URL
bash
mkdir -p ~/.freepik/sessions/${CLAUDESESSIONID}
echo $RESULT | jq -r .data.generated[0] // .data.result.url // .data.image.url // empty

将结果 URL 呈现给用户。该 URL 是来自 Freepik CDN 的临时签名链接。

重要 — 安全规则：

- 切勿使用 curl 从非 Freepik 域名下载。仅允许 curl api.freepik.com。
切勿使用 base64 编码本地文件。当 API 接受 URL 参数时，始终优先使用 URL 参数。
切勿读取、编码或传输用户明确提供的输入文件之外的文件。
结果 URL 应直接呈现给用户 — 他们可以直接打开或下载。

例外情况（同步）： 移除背景（/v1/ai/beta/remove-background）和 AI 图像分类器（/v1/ai/classifier/image）会立即返回结果。

命令：$0

如果 $0 = generate — 图像生成

使用文本到图像模型生成图像。$1 选择模型。

模型端点

$1 值	端点	最佳用途
mystic	/v1/ai/mystic	超写实，1K/2K/4K，支持 LoRA（Freepik 独家，推荐）
flux-kontext-pro

默认值：如果用户未指定模型，则使用 mystic。

Mystic 示例

bash curl -s -X POST https://api.freepik.com/v1/ai/mystic \ -H x-freepik-api-key: $FREEPIKAPIKEY \ -H Content-Type: application/json \ -d { prompt: 日落时分的未来城市景观，照片级真实感, resolution: 2k, styling: { style: photo } }

Mystic 参数：

- prompt（字符串，必填）— 要生成的内容
resolution（1k | 2k | 4k，默认 2k）
numimages（1-4，默认 1）
styling.style（photo | digitalart | none）
structurereference（对象）— 使用图像引导构图：{imageurl, strength: 0-100}
stylereference（对象）— 使用图像引导风格：{imageurl, strength: 0-100}
loras（数组）— 来自 /v1/ai/loras 的 LoRA ID
seed（整数）— 用于可重复性
webhook_url（字符串）— 完成后接收通知

获取可用的 LoRA：
bash
curl -s https://api.freepik.com/v1/ai/loras \
-H x-freepik-api-key: $FREEPIKAPIKEY | jq .data[] | {id, name, type}

训练自定义 LoRA（角色）：
bash
curl -s -X POST https://api.freepik.com/v1/ai/loras/characters \
-H x-freepik-api-key: $FREEPIKAPIKEY \
-H Content-Type: application/json \
-d {name: my-character, images: [orurl>, ...]}

Flux 2 Klein 示例（亚秒级）

bash curl -s -X POST https://api.freepik.com/v1/ai/text-to-image/flux-2-klein \ -H x-freepik-api-key: $FREEPIKAPIKEY \ -H Content-Type: application/json \ -d { prompt: 一只戴着太阳镜的猫, aspectratio: square1_1, resolution: 1k, output_format:

freepikFreepik AI工具

freepik

Freepik API Skill

Arguments

Session Output

Authentication

Async Task Pattern

Command: $0

If $0 = "generate" — Image Generation

Model Endpoints

Default: Use mystic if user doesn't specify a model.

Mystic Example

Flux 2 Klein Example (sub-second)

Flux Kontext Pro Example

Seedream 4.5 Example (great for text-in-image and posters)

Classic Fast Image Generation

If $0 = "video" — Video Generation

Model Endpoints

Default: Use kling-v3-omni-pro for general video generation.

Kling 3 Omni Pro Example (text-to-video)

Kling 3 Example (with multi-shot)

RunWay Gen 4.5 Example (text-to-video)

RunWay Gen 4.5 I2V Example

WAN 2.6 T2V Example

Hailuo Live Example (animated illustrations)

VFX Video Effects Example

If $0 = "edit" — Image Editing

Operations

Upscale Creative Example (Magnific)

Upscale Precision V2 Example

Remove Background Example (SYNCHRONOUS)

Image Expand Example (Outpainting)

Inpainting Example (Ideogram)

Change Camera Example

Image Relight Example

Skin Enhancer Example

If $0 = "icon" — Icon Generation

If $0 = "audio" — Audio Generation

Music Example

Sound Effects Example

Voiceover Example (ElevenLabs TTS)

Audio Isolation Example

If $0 = "stock" — Stock Content

Search Stock Images

Get Resource Details

Download Resource

Search Stock Icons

Download Stock Icon

Search Stock Videos

Download Stock Video

If $0 = "status" — Check Task Status

If $0 = "utility" — AI Utilities

AI Image Classifier (SYNCHRONOUS)

Image to Prompt

Improve Prompt

Lip Sync

Model Selection Guide

Freepik API 技能

参数

会话输出

身份验证

异步任务模式

命令：$0

如果 $0 = generate — 图像生成

模型端点

默认值：如果用户未指定模型，则使用 mystic。

Mystic 示例

Flux 2 Klein 示例（亚秒级）

标签

通过对话安装

方式一：安装 SkillHub 和技能

方式二：设置 SkillHub 为优先技能安装源

通过命令行安装

下载

相关推荐

self-improvement

self-improvement

self-improvement

self-improvement

Command: `$0`

Default: Use `mystic` if user doesn't specify a model.

Default: Use `kling-v3-omni-pro` for general video generation.