OpenClaw Skill · World First

OpenClaw Image Generation & Video Generation: The First AI Skill by Creaa

Let your AI agent search trending content, analyze any product page, then generate marketing images and videos -- all in one conversation. Powered by Nano Banana 2, Seedance 2.0, and 13+ AI models.

We're excited to announce that Creaa.ai is the first AI platform to ship an official OpenClaw skill for AI image generation, image editing, and video generation. If you use OpenClaw as your AI agent framework, you can now access 13+ state-of-the-art AI models directly from your terminal -- no browser, no GUI, no context switching.

But this isn't just "another API wrapper." The real power is what happens when you combine OpenClaw's agent intelligence with Creaa's generation capabilities: your agent can research trending content across the web first, then create optimized visuals in the same conversation. No more switching between tabs, copying prompts, or manually adapting ideas. The agent does the thinking and the creating.

What is OpenClaw? OpenClaw is an open-source AI agent platform that lets you run AI assistants locally. Skills are modular capabilities that agents can invoke -- like generating images, sending messages, or browsing the web. The Creaa AI skill brings professional-grade image and video generation into any OpenClaw agent workflow, and the agent can chain it with web search, data analysis, and other skills automatically.

From terminal to stunning visuals -- the Creaa + OpenClaw workflow

Why Agent + Generation Changes Everything

Traditional AI image and video tools require you to come up with prompts from scratch. With OpenClaw + Creaa, the workflow is fundamentally different: the agent researches first, then creates.

1

Research

Agent browses trending content, analyzes competitors, or reads your product page -- all automatically.

2

Strategize

Agent crafts optimized prompts based on what's actually performing well in your niche.

3

Generate

Agent calls Creaa to generate images or videos using Nano Banana 2, Seedance 2.0, and more.

4

Iterate

Not happy? Tell the agent to adjust. It refines the prompt and regenerates -- no manual editing.

Trending content analysis transforming into polished creative output

Research trending content, then generate -- data-driven creation at its best

Real-World Scenarios

Scenario 1: Trending Content to Viral Video

Tell your agent: "Search for the top trending short-video topics on TikTok this week, pick the best one, and create a 10-second cinematic video." The agent will:

  1. Search the web for current TikTok/Instagram trending topics
  2. Analyze which formats and styles are getting the most engagement
  3. Write an optimized prompt based on what's working
  4. Call Creaa's Seedance 2.0 to generate a 10-second video
  5. Return the video URL, ready to download and post
# Just one message to your OpenClaw agent: Search TikTok and Instagram for the hottest short-video trends this week. Pick the most viral-worthy concept and use /creaa-ai to generate a 10-second Seedance 2.0 video based on it. Aspect ratio 9:16.

Scenario 2: Product Link to Marketing Content

Give the agent a product URL -- Amazon, Shopify, your own store -- and let it handle everything:

  1. Agent visits the product page and extracts key selling points, images, and brand tone
  2. Generates a set of marketing images with Nano Banana 2 (product showcase, lifestyle shots, ad banners)
  3. Creates a product video with Seedance 2.0 (cinematic product reveal, UGC-style demo)
  4. Returns all assets ready for social media, ads, or your store
# Drop a product link and let the agent work: Here's my product: https://www.amazon.com/dp/B0EXAMPLE Analyze this product page. Then: 1. Use /creaa-ai to generate 3 marketing images with nano-banana-2 (1 hero shot, 1 lifestyle scene, 1 social media ad) 2. Generate a 10-second product video with seedance-2.0 showing the product in use
Product page transformed into cinematic marketing content

From product page to marketing video -- one conversation with your agent

Scenario 3: Competitor Analysis to Better Ads

Want to outperform your competitors? Let the agent study their content and create something better:

# Competitive intelligence + content creation in one shot: Search for the top 5 Instagram ads in the skincare niche this month. Analyze their visual style, color palette, and messaging patterns. Then use /creaa-ai with nano-banana-2 to generate 4 ad images that follow the winning patterns but look fresh and unique. Aspect ratio 9:16 for Stories format.

Scenario 4: Blog Post to Video Content

Repurpose your existing content into video without any manual work:

# Turn any article into video content: Read this blog post: https://myblog.com/best-coffee-beans-2026 Extract the 3 most interesting points, then use /creaa-ai to generate a 15-second Seedance 2.0 video for each point. Style: cinematic food photography, warm lighting, slow motion pour shots.

Recommended Models

Our picks: For images, we recommend nano-banana-2 -- best multi-image editing, up to 14 reference images, and excellent prompt understanding. For videos, go with seedance-2.0 -- cinematic quality, flexible 5-15s duration, strong motion consistency.

Image Models

Model Credits Capabilities
nano-banana-2 10 / image Text-to-image, image edit, multi-image edit (up to 14 refs)
nano-banana-pro 20 / image Text-to-image, image edit, multi-image edit (premium quality)
seedream-5.0 4 / image Text-to-image, image edit
gpt-image-2 5-60 / image Text-to-image, image edit, multi-image edit
z-image-turbo 0 / image Text-to-image (fastest & cheapest)

Video Models

Model Credits/sec Duration Capabilities
seedance-2.0 25 5-15s Text-to-video, image-to-video (cinematic quality)
veo-3.1 22 8s Text-to-video, image-to-video (highest quality)
sora-2-pro 12 4-12s Text-to-video, image-to-video
kling-3.0 32 5-15s Text-to-video, image-to-video
hailuo-2.3 25 6-10s Text-to-video, image-to-video
runway-gen-4.5 30 5-10s Image-to-video
sora-2 10 10-15s Text-to-video, image-to-video (most affordable)
Developer workspace with terminal and AI-generated content

Your terminal becomes a creative studio with Creaa + OpenClaw

Installation Guide (3 Minutes)

Setting up the Creaa AI skill takes less than 3 minutes. Here's how:

Step 1: Install the Skill

Choose your preferred method:

Option A: ClawHub (recommended)

npx clawhub install creaa-ai

Or visit the skill page directly: clawhub.ai/yys2024/creaa-ai

Option B: One-line download

mkdir -p ~/.openclaw/skills/creaa-ai && \ curl -o ~/.openclaw/skills/creaa-ai/SKILL.md \ https://raw.githubusercontent.com/yys2024/creaa-ai-skill/main/SKILL.md

Option C: Git clone

git clone https://github.com/yys2024/creaa-ai-skill.git \ ~/.openclaw/skills/creaa-ai

Step 2: Get Your API Key

  1. Go to creaa.ai/profile
  2. Scroll to the API Keys section
  3. Click Create API Key
  4. Copy the key (it starts with ck_) -- it's shown only once

Step 3: Set Your Environment Variable

# Add to your .zshrc or .bashrc export CREAA_API_KEY="ck_your_api_key_here"

Step 4: Verify

openclaw skills list # You should see: creaa-ai Ready

That's it. The Creaa AI skill is now available in your OpenClaw agent. Use the /creaa-ai command or just describe what you want -- the agent knows when to call it.

Why Agent-Powered Creation Wins

  • Research before you create: The agent searches trending topics, analyzes competitors, or reads a product page before generating -- so every output is data-driven, not guesswork.
  • One conversation, full pipeline: From market research to finished content. No more copy-pasting between 5 different tools.
  • Context-aware iteration: The agent remembers your brand, your last request, and your preferences. Say "make it warmer" and it knows exactly what to adjust.
  • 13+ models, zero decision fatigue: Tell the agent what you need; it picks the best model. Or specify Nano Banana 2 / Seedance 2.0 yourself if you prefer.
  • Batch and automate: Generate 20 product images from a spreadsheet of URLs. Create a week's worth of social videos in one session. Build CI/CD pipelines that auto-generate assets.
  • Pay per use: No subscriptions required. Just top up API credits and use what you need.

API Credit Pricing

The Open API uses its own credit pool, separate from your website balance. Credits start from as low as 1 credit per image (Z-Image Turbo) and 10 credits/sec for video (Sora 2). Top up at creaa.ai/pricing.

FAQ

Can OpenClaw generate images?

Yes. With the free Creaa AI skill installed, OpenClaw can generate images from your terminal using 13+ models, including Nano Banana 2. You describe the image and the agent generates it and saves the file for you.

Can OpenClaw generate videos?

Yes. The same Creaa skill adds text-to-video and image-to-video to OpenClaw, powered by models such as Seedance 2.0 and Sora 2 -- all triggered with the /creaa-ai command.

Is there an OpenClaw image generation skill?

Yes -- the Creaa AI skill was the first OpenClaw plugin for image and video generation. Install it via ClawHub (npx clawhub install creaa-ai) in about 3 minutes.

Do I need an OpenClaw account?

You need OpenClaw installed locally. The skill itself is free. You only pay for Creaa API credits when generating content.

What if a generation fails?

If a task fails before it starts processing, credits are refunded automatically. If a queued task fails, credits are also refunded to your API key balance.

Can the agent really search the web and then generate?

Yes. OpenClaw agents can chain multiple skills in a single conversation. The agent uses its built-in web browsing to research, then calls the Creaa skill to generate. You just describe the end goal.

Which model should I start with?

For images: nano-banana-2 (best all-rounder with multi-image editing). z-image-turbo is a low-cost option at 1 credit per image. For videos: seedance-2.0 (cinematic quality, flexible duration). For budget: sora-2 (10 credits/sec).

Can I give it a product URL and get content automatically?

Absolutely. Paste any product link -- Amazon, Shopify, Etsy, your own store. The agent reads the page, understands the product, and generates marketing images and videos tailored to it.

The Creaa AI skill is open source and available on GitHub. Contributions and feedback are welcome.
OpenClaw is a trademark of its respective owners. Creaa.ai is not affiliated with the OpenClaw project.