Not sure what to post this week? OddPen follows creators in your field, shows you why their videos went viral, and provides you with scripts tailored to your own style that you can shoot right away — going from a blank page to filming in just a few minutes, not days.
Three steps — from a channel you follow to a script you can film today.
Drop in the YouTube channels and podcasts you already follow, plus any Substack newsletters. While the app is open, OddPen watches them and pulls in every new post for you — ready to transcribe and break down in one click.
OddPen pulls each post apart and tells you, in plain words, why it took off — the opening that hooked people, what kept them watching, and the angle nobody has covered yet.
Out come three fresh angles and full scripts, ready to film — written to sound like you, not like generic AI.
Most AI tools just spit out generic content and hope it sticks. OddPen does the opposite: it reads everything in your niche, then explains — in plain English — exactly why a video took off. The line that hooked people in the first three seconds, the moment they couldn't look away, and the angle every other creator missed. You stop guessing, and start from what already works.
Opens with a bold claim in the first three seconds.
Builds tension, takes a turn, then releases.
The angle they left wide open for you.
Generate three fresh angles and ready-to-shoot scripts — written in your persona's style, not generic AI filler. Set your tone once, reuse it forever.
Everyone else rents you AI by the month — and quietly marks up every word it writes. OddPen doesn't. You buy it once, run it on your own computer, and plug in your own AI key from Google, OpenAI or others — most have a free tier, so it can cost you almost nothing to run. No monthly bill, no markup, and nothing you research or write ever leaves your machine.
If you publish regularly and "what do I make this week?" is the hard part, OddPen is for you.
Turn the channels you follow into a steady stream of angles and ready-to-film scripts.
Turn episodes into clean transcripts, and see exactly what makes them land with listeners.
Spot what your audience is into, then turn it into your next lesson or post.
Find what's worth writing about, see why it resonated, and start your draft with a real angle — never a blank page.
You buy OddPen once and run it on your own computer. Here's how that stacks up against renting a tool month after month.
An illustrative look at how OddPen works.
Opens with a bold claim in the first three seconds.
Builds tension, takes a turn, then releases.
The angle they left wide open for you.
Opens with a bold claim in the first three seconds.
Builds tension, takes a turn, then releases.
The angle they left wide open for you.
One desktop app, no add-ons.
Turn any video or audio into clean text. Local Whisper even handles videos with no captions.
AI breakdowns of every viral piece you feed it.
Three fresh takes plus full scripts per topic.
Scripts written in your persona's own style.
Plug in a free LLM key — no token markup, ever.
Works offline; your research stays yours.
What it runs on, the models it talks to, and what you can tune.
Windows 10 / 11 desktop app
macOS — on the roadmap
No GPU needed — about 8GB RAM recommended
Free tier: Gemini, Qwen, Groq
Paid: DeepSeek, OpenAI, Claude
Sources: YouTube, podcasts, Substack
Local Whisper transcribes videos with no captions
Checks your sources every few hours while the app is open
Saves & exports to a folder you choose
OddPen makes about three AI calls per piece — the breakdown, the three angles, and your script — plus optional chat to refine the script. That’s roughly 12k tokens per video, billed straight to your own key with zero markup from us.
What 1M tokens gets you (≈ 750k words): about 75 full runs — transcribe → breakdown → 3 angles → script — or a few hundred refine-the-script chats.
Free tiers carry daily / per-minute limits and can run out — past them, paid keys are billed directly by the provider at their own (low) rates. Per-video token use is an estimate; your costs scale with how much you analyze. Prices change often, so check each provider’s official pricing: Groq · DeepSeek · Gemini · OpenAI · Anthropic.
Buy once. No subscription, ever.
Windows 10 / 11 desktop app.
No GPU required.
Internet only for fetching content & your LLM calls.
Your own LLM API key (a free tier works fine).
One license = up to 2 computers.
Buy once. Updates included, no recurring fee.
Buy once and every update to this version is included.
No — OddPen is a one-time purchase. You own this version forever and get its updates. No monthly fee, ever.
It can be, for light use. Gemini, Groq and DeepSeek all have free tiers, but those come with daily and per-minute limits and can run out. Past them, a cheap paid key runs about a penny per video — and we never add a markup, since you pay the provider directly.
Bring your own key for OpenAI, Anthropic Claude, Google Gemini, Groq, DeepSeek or Qwen. You pick the one that fits your budget.
So the AI cost stays between you and the provider, with no markup from us. Start on a free key; switch to a paid one only if you outgrow it.
OddPen falls back to local Whisper speech-to-text, so it can still transcribe videos that don’t publish captions — right on your computer.
YouTube channels, podcasts and Substack newsletters. While the app is open, OddPen checks them for new posts every few hours.
While the app is open, OddPen checks your sources on a schedule (every few hours) and pulls in new posts — so your feed is ready when you look. It only fetches in the background; transcribing and analyzing happen when you click an item, so it never runs up your AI key on its own. Close the app and the schedule just pauses.
A Windows 10 / 11 desktop app today; macOS is on the roadmap. No GPU required (a GPU only speeds up local Whisper). One license activates on up to 2 computers.
Only the content you ask it to fetch and the prompts sent to your chosen LLM. Everything else — your sources, transcripts and scripts — stays local.
If it’s not the right fit, see the refund policy for details.