Ask what a faceless YouTube channel costs to run and you'll get two kinds of answer. One says it's free. The other quotes a stack of subscriptions that adds up to more than most people's car payment.
Both are useless, because neither tells you the only number that matters: how many views a month you need before this stops costing you money.
Here's the honest version.
The costs that are real
Prices move constantly, so treat these as bands rather than quotes, and check current pricing yourself before committing to anything.
The engine — unavoidable
If you're running AI-assisted production, you need a paid plan on whatever model you're using. Free tiers will not carry real workloads; you'll hit limits mid-run and start making decisions based on what you have quota for, which is a terrible way to choose a topic. Budget for one paid plan. This is the one line item you cannot talk yourself out of.
Video assembly — unavoidable
Something has to turn a script into a finished video: voice, footage, edit. Whether that's a dedicated faceless-video tool or a stack you assemble yourself, there's a monthly cost, and it scales with how much you publish.
Thumbnails — optional, has a free route
Dedicated thumbnail tools are fast and good. A free image generator plus disciplined text-setting gets you most of the way for nothing. Start free. Upgrade when the channel is earning and the time saved is worth more than the subscription.
Market data — optional, but this is where it hurts
Channel-level data on revenue, growth and outliers is the difference between choosing a niche with evidence and choosing one with vibes. It's the most skippable subscription and the one whose absence costs the most, because it doesn't hurt in month one. It hurts in month eight when you find out the niche was never viable.
Everything else
Email, a domain, a community platform, design tools. All real, all deferrable. None of them belong in month one.
The cost nobody puts in the spreadsheet
Your time is the largest single expense in this business and it never appears in anyone's cost breakdown.
Research, topic selection, scripting, packaging, thumbnails, uploading, checking the analytics. Even heavily systematised, a single video is hours of real work. At any honest hourly rate, the human time on a faceless channel dwarfs every subscription combined — which is exactly why "just upload more" is expensive advice, and why the tooling question is really a time question wearing a cost costume.
The only number that matters
Take your total monthly running cost. Divide it by your RPM — your revenue per thousand views, which you'll find in your own analytics rather than in someone's video. Multiply by a thousand.
That's your break-even in views per month.
Do this before you start and something clarifies immediately: the niche's RPM changes the answer more than any subscription you cancel. A low-RPM niche can need several times the views of a high-RPM one to clear exactly the same costs. Which means the single biggest cost decision you'll ever make isn't which tools you buy — it's which niche you enter. That decision is made before you spend a dollar, and it's almost impossible to undo later.
What we'd actually tell you to do
Start narrow. Pay for the engine and the video assembly, take the free route on thumbnails, and validate the niche before you add anything else. Write your break-even number down where you'll see it.
Then let the channel earn its own tooling. Every subscription you add before the channel is earning is a subscription funded by optimism, and optimism has a much shorter runway than most people expect.
If you want the whole list of what we pay for and what we skip, it's on our tools page — including the ones where we have no affiliate arrangement at all.