Drop in your avatar, set a batch size, confirm the cost — Claude does the rest. Real face preservation. Amateur iPhone aesthetic. $0.04/image.
One Claude skill that turns an avatar photo into an entire content library — selfies, mirror shots, varied outfits, locations, and emotions — all automatically.
Seedream 4.5 Edit is a true image-to-image model. It reads your reference photo and keeps the face, skin tone, and proportions across every single generation.
Every prompt is engineered to look like a real candid selfie — raw framing, natural imperfections, no studio lighting, no editorial vibes.
Claude generates all prompts, writes the batch script, launches it in the background, and delivers the finished photos. You just say how many.
If a batch fails halfway, relaunch it — it skips already-completed images and picks up where it left off. Never pays twice.
One-time setup, then it's just "generate X photos" every time after.
Download from python.org — check "Add to PATH" during install. Then open PowerShell and run pip install requests.
Download the zip below. Extract the wavespeed-batch-api folder and drop it into your Claude skills directory:
C:\Users\YourName\AppData\Roaming\Claude\skills\
Restart Claude so it picks up the new skill.
Sign up at wavespeed.ai, grab your API key from the dashboard, and top up your balance. Every image costs $0.04.
Upload your reference photo to imgbb.com (free). Copy the Direct Link — it looks like https://i.ibb.co/xxxxx/photo.jpg. WaveSpeed pulls from this URL directly.
Claude walks you through API key, avatar URL, and identity setup. Takes about 2 minutes. After that, just say generate 50 photos and let it run.
Save your identity file. After first setup, Claude saves a wavespeed_identity_name.md file. Add it to your Claude project and skip onboarding forever.
Use the direct image URL. imgbb gives you two links — use the one starting with i.ibb.co, not the viewer page. WaveSpeed needs the raw file URL.
50/50 selfies to mirror selfies. Claude enforces this automatically — variety keeps the content pack looking natural and diverse.
Failed batch? Just say "relaunch." It checks the checkpoint, skips what's already done, and picks up from where it stopped. No duplicate charges.
Custom sizes. Leave size blank for model default, or specify dimensions like 1024x1792 for portrait-optimized output. Supports up to 8192×8192.
Boost your workflow with these copy-ready prompts. Drop them straight into Claude to reverse-engineer images or analyze OnlyFans dynamics.
Download the skills, client scripts, and setup guides to deploy these automated pipelines inside Claude.
Claude skill and Python client to generate candidate AI influencer images ($0.04/image) with face preservation and raw iPhone aesthetic.
Claude skill that automates short-form video clips generation with custom text overlay burned in via FFmpeg (~$0.32/clip).
Complete guide detailing prerequisite installs (Python, FFmpeg) and deployment steps for the TikTok video pipeline.
This tool is just one piece. The AI Creators Club gives you the full blueprint — models, workflows, automation stacks, and a community of people actually doing it.
aiofmblueprint.com