Frameghost

You want a photo of yourself looking like you were shot on a disposable camera in 1998. Or leaning against a wall in Tokyo at golden hour. Or a CCTV still, slightly grainy, like a scene from a film.

You describe it to an AI image generator and get something that looks like a stock photo.

The problem isn't the AI. It's the prompt. Generating a realistic photo of yourself requires vocabulary most people don't have — the right film stock, lens focal length, lighting angle, compositional framing, outfit description, atmospheric conditions. Professional photographers know this language. Most people don't, and learning it just to describe yourself is not a reasonable ask.

Frameghost handles the language. You handle the vision.

Here's how it works: Frameghost opens as a side panel inside Chrome, right next to your AI image generator — ChatGPT, Gemini, Grok, or any other. You pick from a set of visual options — what film style, what scene preset, what outfit, what pose, where, what time of day, what mood. As you pick, it builds a prompt in the background. When you're done, hit Copy and paste it in. That's it.

What you're picking from matters. There are two distinct style layers. The Film & Camera picker covers photographic medium — Polaroid, disposable camera, Kodak Portra, VHS, Super 8, CCTV, slide film, infrared — each translated into the exact grain structure, color cast, and tonal detail the AI needs to render it. You're not typing the word vintage. You're selecting what vintage means photographically, and Frameghost writes that out for you.

The Scene Preset picker is something different: complete atmosphere setups. Neon Street. Bedroom Lo-fi. LCD TV Reflection. Tatami Room Light. Rooftop Golden Hour. These aren't filters — they're fully written environmental contexts. Select one and Frameghost automatically sets the location, time, and atmosphere together as a coherent scene. You can still layer a film stock on top.

Same goes for everything else: outfit, pose, expression, composition, location, weather, mood. Location goes deep — beyond "café" or "alley", you can specify a Tokyo konbini, a Seoul back alley, a Paris Métro platform, a Kowloon rooftop with illegal structure additions and laundry lines. 8 countries, each with multiple settings that tell the AI exactly what kind of place to render. You pick. It writes. The result is a prompt that sounds like it came from someone who spent years behind a camera — and has been to these places.

There's nothing to learn. You either know the look you want or you explore until something clicks.

Free tier includes 15 film styles and core options. One-time $7.99 unlocks everything: the full Film & Camera library, all 34 Scene Presets (atmosphere setups like Neon Street, Bedroom Lo-fi, Tatami Room Light, Cybercafe Glow), all outfit, pose, expression, and composition options, a vintage date stamp for any month from 1960 to 2030, and cultural location settings for 8 countries — Japan, Korea, Taiwan, Hong Kong, Morocco, France, Thailand, and the UK. Each country has hand-crafted location fragments: Tokyo konbini, Seoul subway, Paris pavement café, Kowloon rooftop, Marrakech medina alley — specific enough that the AI knows exactly what to render.

No subscription. No account. Works with any AI image generator — ChatGPT, Gemini, Grok, and more.

If your AI photos don't look like you imagined, it's almost always the prompt.

—

30-day refund guarantee — not happy? Email us within 30 days.

PRIVACY: No browsing data collected. Selections are saved locally only. A device ID is used solely to validate your license key — never shared or sold.