the official Fanvue AI creator program* πŸ““ Workbook
πŸ”’ Heads up β€” finish the previous lesson first. The course works in order.

Phase 2 · First Publish · ⏱ 15 min

Your First AI Photoshoot

β–Ά {{VIDEO: lesson 2-1 β€” Owen & Cole}}Video drops in here. Everything below stands on its own.

When you're done

Your creator has her first real content: a four-photo set, shot in the studio, sitting in your Library ready to publish. First shoot takes fifteen minutes. Every one after takes five.

Why this matters

Content is the product. Everything downstream β€” the profile, the posts, the subs, the paid unlocks β€” is made of what you generate right here. And here's the unlock most people miss: one shoot gives you a set β€” profile photo, banner, feed post, and a social post, all consistent, all from the same fifteen minutes. You're not making a picture. You're stocking inventory.

The steps

Watch the whole shoot happen once β€” then do yours below.
Watch the whole shoot happen once β€” then do yours below.
  1. Open the Create tab. Quick orientation: modes across the top (Single / Carousel / Scene swap) and the rating toggle (SFW / NSFW). Today: Single, SFW.

SFW today, and in Bopgpt until Phase 6 β€” on purpose. (The Premium Studio section has its own rules when you get there.) Traffic platforms need a clean feed, and the money logic for spicier content lives where the money lessons live. Carousel gets its own lesson in Phase 3; Scene swap lives in the Vault.

  1. Confirm your model β€” the panel on the right shows who's shooting. Should be your creator; switch if it isn't.

  2. Build the prompt from cards β€” one per category. Here's the first-shoot recipe. Run it exactly:

Category Card
Outfit Casual
Setting Bedroom (or Beach if her vibe says so)
Pose Selfie
Lighting Soft natural
Camera iPhone

Why this recipe? It reads like a real person's first post β€” casual and candid, not a magazine ad. Authentic beats polished for a brand-new account, every time.

  1. The text box β€” one detail, optional. The cards already wrote the prompt; this box is seasoning, not the meal.

smiling, holding iced coffee

  1. Settings on the right: Images 4 Β· Aspect Portrait (it's a phone-first world) Β· Quality: leave the default Β· Realism Ultra-real β€” that's the "reads like a real phone photo" switch. Polished is for glam sets later.
The recipe, loaded: five chips, one seasoning line, 4 Β· Portrait Β· Ultra-real.
The recipe, loaded: five chips, one seasoning line, 4 Β· Portrait Β· Ultra-real.
  1. Generate. A set like this runs 5 credits per image β€” 20 for the set. {{COPY: your Lab plan's included credits β€” pending #41}}. Watch the four results land.
Lena's actual first set β€” same face four times, coffee in hand, reads like a real Tuesday. This recipe gave 4/4 keepers.
Lena's actual first set β€” same face four times, coffee in hand, reads like a real Tuesday. This recipe gave 4/4 keepers.
  1. Pick keepers with the 10-second rule. Per photo: hands normal? signature trait visible? background clean? face consistent with her reference? Expect 3–4 keepers from a good recipe β€” one weird one is normal, that's why we shoot four. Fewer than 3 keepers? Run the same recipe once more and take the best across both sets. (Aim for 4 β€” the next three lessons use these for her avatar, banner, first post, and first social post, and reusing a shot between the banner and a social post is completely fine.)

  2. Open the Library and download the best one to your device. That's tomorrow's profile photo.

AI prompts for this lesson

Copy these into your AI assistant when you want a hand. They know this lesson.

Not sure which shot is "the one"? Get a second opinion:

I just shot my AI creator's first content set. Here's her persona: [paste sheet]. I'm choosing her profile photo. For each of my favorites I'll describe them: [photo 1: ..., photo 2: ...]. Which reads most like a real person's best casual photo β€” and why, in one sentence? Profile photos need: face clear at thumbnail size, warm, believable.

Do it now

~15 minutes.

You're done when

A generated set from your model sits in your Library with 3+ keepers, and your best image is downloaded. This is the moment the business stopped being an idea β€” celebrate it in #content. We mean it. Every single first shoot gets hyped in there.

If you're stuck

The photos don't look like my model

Open Models β†’ your model β€” is the reference image right? If the model's fine but one photo drifted, that's normal variance: keep the consistent ones. If ALL four drift, regenerate once; still wrong β†’ #bopgpt-help, something's off with the reference and that's ours to fix.

Hands or details look off in one photo

Delete it, keep the rest. One-in-four is the AI tax. Never publish the weird one β€” your keeper rate is the quality bar fans see. (And deleted means deleted from your Library β€” nothing you cull goes anywhere. {{VERIFY: deletion behavior at filming}} Nothing publishes without you; the studio's filters plus your cull are the two-lock system.)

Generation failed or says content flagged

The safety filter tripped. Two causes, in order: wording in the text box (clear it, regenerate with cards only), or an uploaded reference it didn't like (remove it, retry). Repeats β†’ reword the detail; no brand names, no real-person references.

I ran out of credits

A 4-image set costs 20 credits. {{COPY: per-plan credit guidance + top-up flow β€” pending #41}}.

Can I try other cards now? What about NSFW?

Other cards: {{COPY: spare-set guidance depends on your plan's credit allocation β€” pending #41}}. NSFW in Bopgpt: not until Phase 6, deliberately (the Premium Studio covers its own lane later). Clean feed first; money logic later, where it earns.

Next up

You've got inventory. Next lesson we turn the best of it into her storefront β€” avatar, banner, bio, and her subscription price, live on Fanvue.