All my free AI prompts, guides & systems
Real builds, real workflows. Every guide has the exact prompts, the exact commands, and the honest caveats — so you can replicate the result, not just watch it.
The complete ChatGPT guide
Every plan, every model, every feature. Verified 22 September 2026.
Every plan, every model, every feature that is live right now — what each one costs, and what it is still bad at. No fluff, no affiliate links, nothing behind an email form.
The complete Claude guide
Every plan, model and product. Verified 23 September 2026.
Every plan, model and product that is live right now — what each costs, and what Claude still cannot do. No affiliate links and no email form.
Showing 16 guides
Build a $5,000 AI receptionist in a weekend
Claude Code × Fish Audio
Claude Code writes the entire app from one prompt. A free local model runs her brain. Fish Audio is her voice. One API key, one weekend.
Build a faceless channel that doesn't sound AI
Claude × Fish Audio
Claude writes the script as a performance — emotion tags dropped inline. Fish Audio acts it out. Two tools, the whole channel's audio pipeline.
Find your first customers with an AI agent
Runable
One agent that works out who your customer actually is, finds them, writes to them from your own inbox, and chases the follow-ups you'd never keep up with.
Run a faceless channel on autopilot
Claude × FacelessReels
Claude picks the niche that actually pays. FacelessReels builds and posts the reels daily while you sleep. Plus the step almost every autopilot guide skips — the one that decides whether your channel earns or gets demonetised.
From "no idea" to a real business in one afternoon
Replit Free Mode
The website, the app, the designs, the deck, the competitor analysis, the GTM report, a hundred prospects and the emails to send them — one afternoon in Replit. Here's the whole chain.
Build your own JARVIS
Claude × Fish Audio
The assistant that tells you what's actually on fire. Out loud. Claude reads your real apps, judges by your rules, Fish Audio reads it back to you. The real build, including the failures.
Stop editing your talking-head videos manually
OpusClip AI Producer
One raw take, no timeline. Drop the file, brief it in plain English, and ten minutes later it's done — ums gone, b-roll placed, captions styled, motion graphics on the beats. The whole workflow.
Catch a trend while it's still worth catching
Medeo VideoClaw
A live feed of what's accelerating right now, a breakdown of why it worked, and Medeo to turn it into your video. The full loop — find, understand, make, ship, learn.
Design any character voice — and then direct it
BreezeBlue, Breeze TTS 2
The soccer commentator, the gravelly documentary narrator, the tiny chirpy bird — none of those voices existed until I typed them. This is the full prompt library: how to design the voice, then how to direct the performance.
Design any voice without writing the prompt yourself
Claude × Fish Audio
The soccer commentator, the children's storyteller, the tiny chirpy bird — all back in one conversation, without leaving Claude, without browsing a voice library, and without writing a single detailed voice description yourself. That last part is the actual trick.
ChatGPT visual codes
One word. One product photo. Hit enter.
Every visual in that carousel came out of ChatGPT — one word, one product photo, hit enter. Nothing to install, nothing to set up. Here are the codes I used, plus fifty more.
Write a scene, cast it, hear it
GPT-6 Astra × Fish Audio
Two characters, a full conversation, two different voices — and never a line of dialogue or a single voice description written by hand. GPT-6 Astra writes the scene and tags the emotion into it; Fish Audio performs each character in their own voice.
100 ChatGPT AI image codes
Ten categories. One line. Any photo.
Type one line into ChatGPT and turn any photo into a completely different creative direction — exploded views, x-rays, blueprints, anime, product ads, and more. One hundred codes, ten categories, no prompt writing.
100 AI video codes
Google Flow — still photo to motion in one line
Type one line into Google Flow and turn any still photo into motion — camera moves, product spins, elemental FX, talking avatars, and more. One hundred codes, ten categories, no prompt writing.
200 ChatGPT image codes
Thirteen detailed codes + the full two-hundred library.
Thirteen codes I showed in the carousel with the full prompt behind each one, then the whole 200-code library across 14 categories. One code each, no prompt writing.
Edit your talking-head videos with Codex
Raw clip in, finished vertical reel out. No editing software.
Drop a raw talking-head clip into a folder, paste one prompt, and Codex hands back a finished vertical reel: silences cut, captions timed word by word, key lines punched in with a zoom and a headline. Setup takes one afternoon. After that, ten minutes per video.