AI Video Cheat Sheet October 2026: Everything You'll Ever Need
Every tool I use to plan, record, edit, generate and voice video, with what each one is actually good for and a link to start. Includes the full tier list from the video.
Phone is optional. By adding it you agree Orgrowth may call or text you about this; message and data rates may apply, and you can opt out any time.
You'll also receive my content strategy newsletter with more free breakdowns.
Every tool, in the order you'd use it
Claude Code
The middle of the whole system. Claude Code runs on your computer, reads and writes your files, and plugs into every other tool on this list through MCP connections, so one chat can research, edit and generate your video.
- ▸Connect Notion, Higgsfield and Palmier Pro so Claude can read your knowledge base and drive each tool for you
- ▸Save your best prompts as skills, like a prompt writer for Seedance or Kling, so you get the same quality every time
- ▸Download, transcribe and break down reference videos so it knows what is working in your niche
- ▸Run FFmpeg and Gemini for you, so it can actually watch and listen to your footage
Notion
Where my Brand Brain lives: my ICP, offer, goals and content formats. Every AI tool reads from it, which is the difference between your voice and generic AI slop.
- ▸Keep one page with your audience, offer and voice that every prompt reads first
- ▸Run your content calendar and script database in one place
- ▸Connect it to Claude so it can read your knowledge base and write back to it
yt-dlp
Free, open-source downloader for YouTube, Instagram and thousands of other sites. Claude runs it for you, so any reference you paste becomes a file it can actually study.
- ▸Download reference reels and YouTube videos to break down shot by shot
- ▸Pull just the audio, ready to transcribe with Gemini
- ▸Grab captions, view counts and upload dates without downloading the whole video
- ▸Save the VSLs and funnel videos you want to learn from
FFmpeg
Free command line software that handles video. On its own it's intimidating; with Claude driving it, it's how Claude actually sees your footage.
- ▸Pull still frames from any video so Claude can look at the shots
- ▸Cut, trim, join and convert clips without opening an editor
- ▸Check loudness and fix audio levels before you post
Gemini
Google's AI model. I use it through the API to transcribe video and audio, because it's cheap and handles long files well.
- ▸Transcribe your reels and reference videos with timestamps
- ▸Summarise long videos and podcasts before you decide to watch them
- ▸Get an API key in Google AI Studio and let Claude run it for you
Riverside
The best way to record calls and interviews. It records 4K locally on both ends, so a bad connection never ruins the footage.
- ▸Record podcasts, interviews and client calls in 4K from both devices
- ▸Zero lag in the final file, even if the live call stuttered
- ▸Turn long recordings into short clips with Magic Clips
Tella
My pick for screen recordings. No lag, and it does the boring cleanup for you.
- ▸Record your screen and face together without lag
- ▸Automatically cut the silences out of your recording
- ▸Export the screen and your face separately so you can edit them as two layers
OBS
Free and reliable, but clunky. It's the backup when you don't want to pay for a recorder.
- ▸Record your screen and camera for free
- ▸Stream live to YouTube and other platforms
- ▸Expect some setup: separating the screen from the face cam afterwards is hard
Loom
Fine for a quick message to a teammate, but not for content. The video gets choppy if your internet isn't amazing.
- ▸Send quick walkthrough videos to clients or your team
- ▸Share by link instantly with no export
- ▸Skip it for anything you plan to post
Palmier Pro
A full video editor built for AI. Claude drives the timeline through MCP, and you can still adjust everything yourself like you would in CapCut.
- ▸Edit your video from one prompt in Claude, then fine-tune it on a normal timeline
- ▸Cut silences, add captions and rearrange clips by describing what you want
- ▸Generate images, video and audio right inside the project
- ▸The editor and the MCP connection are free; you only pay for generation credits
CapCut
Super quick for yap videos, and the captions are goated.
- ▸Cut a talking head video fast on your phone or desktop
- ▸Add auto captions that look good straight away
- ▸Use templates and effects without learning a pro editor
Instagram Edits
Meta's free editing app. The captions are limited compared to CapCut, but I use it to give the Instagram algorithm a little nudge.
- ▸Do your final pass in Edits before posting to Instagram
- ▸Export in 4K with no watermark, and record clips up to 10 minutes long in the app
- ▸Use the built-in voice enhancement and captions for quick fixes
Descript
Great concept, editing video by editing the text, but it's super clunky to use.
- ▸Cut video by deleting words from the transcript
- ▸Remove filler words in one click
- ▸Worth a try if you hate timelines, but expect it to feel slow
Premiere Pro
Adobe's a bit of a scam for most creators. Your editor should use this for complex stuff, but you probably don't need it.
- ▸Hand complex edits and long-form projects to an editor who lives in it
- ▸Use it when you need multicam, detailed audio work or tight colour control
- ▸Skip the subscription if you're editing your own short-form
HyperFrames
HeyGen's open-source video tool that AI agents can write. Claude builds the captions and graphics as code, so you describe the look and it renders.
- ▸Add word-by-word captions synced to your voice
- ▸Build animated titles, charts and motion graphics from a prompt
- ▸Install it into Claude Code in one step (the setup is in my guide below) and keep every render on your own machine
Remotion
Make videos with code. Claude writes the composition, Remotion renders it, which is perfect for captions and text you want identical every time.
- ▸Render branded captions and text overlays that match your style every time
- ▸Batch render many versions of one video from a template
- ▸Free for individuals and small teams; check their licence if you're a bigger company
Colourlab AI
AI colour grading. Colourlab AI 4 balances and matches your shots so footage from different cameras looks like one film. Version 4 was in free public beta when I wrote this.
- ▸Balance every shot in a scene in one click
- ▸Match footage from your phone, DJI and main camera so they cut together
- ▸Apply cinematic film emulation looks
GPT Image 2
OpenAI's image model, the one behind ChatGPT Images 2.0. My go-to for thumbnails, start frames and anything with text in it.
- ▸Generate images from a text prompt inside ChatGPT
- ▸Edit images you upload, like swapping a background or an outfit
- ▸Create start frames to animate in Kling or Seedance
Kling 3.0
Turns a still image into a moving shot. Kling's 3.0 models are where I go when I need believable motion from one frame.
- ▸Animate a still image into a video clip
- ▸Copy the motion from a reference video onto your character with Motion Control
- ▸Generate in native 4K
Seedance 2.0
ByteDance's video model. It handles several scenes in one generation, which makes it the best option for short cinematic sequences.
- ▸Generate a multi-scene clip from one detailed prompt
- ▸Start from your own image so the character stays consistent
- ▸Use it inside Higgsfield or invideo instead of signing up separately
Higgsfield
Every major image and video model in one place. The MCP lets Claude generate for you, and Genjutsu Restyle turns real footage into an animation style.
- ▸Connect the Higgsfield MCP so Claude generates images and video straight from your chat
- ▸Restyle any video into a new animation style with Genjutsu Restyle, keeping the original motion
- ▸Swap objects or transfer motion into a new scene while keeping the rest of the shot
invideo Agent One (now Agent Two)
An AI video agent you brief like a film crew. It remembers your characters, locations and style across every shot, and picks the right model for each one.
- ▸Turn a script into a storyboard, shots, voices and music
- ▸Keep characters consistent across a whole film
- ▸Change one thing everywhere at once, like flipping every scene from day to night
- ▸Final assembly still happens in your editor
VidRush
A long-form video generator built for faceless YouTube. You give it one brief and it hands back a full video ready to edit and publish.
- ▸Turn one brief into a long-form faceless YouTube video
- ▸Get script, visuals and edit in one place
- ▸Use it to test a faceless channel idea before hiring an editor
Fish Audio
AI voices that sound natural, with emotion tags you write straight into the script. It's what I use for faceless voiceovers.
- ▸Clone your own voice from a short recording
- ▸Steer the delivery with tags like [relaxed] or [excited] in front of a line
- ▸Pick from a huge library of community voices in different accents
ElevenLabs
The best known AI voice platform, with polished voices and more than just text to speech.
- ▸Generate voiceovers from a script
- ▸Clone your voice for consistent narration
- ▸Create sound effects and dub videos into other languages
HeyGen
Lip syncs a voiceover onto video, so a new script matches the mouth on screen.
- ▸Sync a new voiceover to existing footage
- ▸Translate a video into another language with matching lip sync
- ▸Create an AI avatar of yourself for quick talking videos
You
Final approval, still youAI does the 90: the research, the first drafts, the edits and the generation. You do the 10 that matters: the last pass on every draft, your face on camera, and the call on what's worth posting. Take yourself out and you're one more faceless account competing with ten thousand others.