Last month an agency owner showed me two invoices. One was for a freelancer who charged $400 for a 45-second reel and took nineteen days. The other was ours: $70, delivered in two. He asked the question everyone asks at that moment, what's the catch?
This post is the honest answer, route by route. No "AI is the future" filler. Just what each option actually costs per finished video in 2026, including the costs that never make it onto the quote. A warning before the numbers: pricing in this market moves fast, and any figure you read anywhere, including here, has a shelf life. What changes slower is the structure underneath the prices, where each route hides its real cost. That structure is what this post is really about, because the structure is what lets you evaluate any quote you'll ever receive.
Route 1: DIY tools, $20 to $100 a month, plus your nights
The subscription looks harmless. Twenty, fifty, a hundred dollars a month for text-to-video, AI voice, captions. The marketing implies you'll knock out videos over coffee.
The reality: your first usable 60-second reel takes four to eight hours. Not because the tools are bad, but because getting generated footage to match your brand takes iteration, and the edit is where the real time goes. At any professional hourly rate, your "cheap" video costs $150 to $500 of your own time, and it still looks like the learning curve, because it is.
There's a second cost nobody budgets for: tool sprawl. Generation, voice, captions and music rarely live in one subscription, so the real DIY stack is three or four subscriptions and the glue work between them. Export from one tool, upload to another, re-check the caption timing in a third. That glue work is pure overhead, and it doesn't shrink with practice the way generation skill does.
Where DIY makes sense: testing an idea before commissioning it, personal brands where the founder's time is genuinely free, and learning what the technology can and can't do so you can brief a production partner properly later. That last one is underrated. An owner who has spent ten hours inside these tools writes briefs that save a production partner hours, and better briefs show up as better videos at the same price.
Where it doesn't: any consistent publishing commitment. A daily posting rhythm dies the third week, when the novelty wears off and the backlog starts. I've watched it happen to enough teams that I now treat "we'll do it in-house with tools" as a three-week experiment, not a plan.
Route 2: Freelance editors, $150 to $800 per video
A good short-form editor charges $150 to $500 per reel in 2026, more with motion graphics. The rate is fair for the work. The problem is everything around the rate.
Good freelancers are booked two to three weeks out. Your rush job sits in a queue behind someone else's retainer. Revisions go through rounds of messages. And when they take a holiday, get sick, or land a bigger client, your content calendar stops. None of this is a criticism of freelancers; it's the shape of the arrangement. One person's calendar is a single point of failure, and your publishing schedule inherits that risk whether you've acknowledged it or not.
There's also management time, writing briefs, chasing files, checking versions. Add two to three hours per video of your own coordination and the true cost per video lands between $200 and $900 depending on volume and how much you chase. The coordination cost is sneaky because it never appears on an invoice, but it's real payroll: it's hours you or a team member spend being a producer, whether or not either of you is good at it.
Where freelancers make sense: low-volume, high-touch projects. Brand films. Anything where the back-and-forth is worth it. Not for daily content. A useful test: if you need more than eight videos a month, you don't have a freelancing relationship, you have an under-resourced department.
Route 3: In-house editor, $45,000 to $95,000 a year
A full-time short-form editor in the US or UK costs $45k to $95k plus tools, plus management, plus the office chair. Capacity: 15 to 25 videos a month if they're organized.
Run the per-video math honestly: $175 to $450 per video, before salary inflation, sick days, and the day they resign with all your brand knowledge in their head. If they're good, they'll leave for a higher offer within eighteen months. If they're average, you'll know by month three and start again. Either way you're re-hiring on a cycle you didn't plan for, and re-hiring is the expensive part: recruiting, onboarding, and two or three months of output that isn't quite on brand while the new person learns your standards. Budget the re-hire, not just the hire.
There's also the utilization problem. A salaried editor between projects is still being paid, so the effective per-video cost swings with how well your pipeline keeps them fed. Most teams overestimate their content appetite when they hire, then feel guilty assigning the editor to slide decks to keep them busy. And the hiring market has its own motion: good short-form editors know their worth in 2026, counteroffers are routine, and the salary you budgeted in January rarely survives the first competing offer in September.
Where in-house makes sense: large brands with steady volume, deep brand systems, and work that's genuinely confidential. For everyone else it's the most expensive way to buy 20 videos a month.
Route 4: AI production partner, $70 to $150 per video
This is what NNI does, so weigh my bias accordingly. A production partner runs the full pipeline: AI-generated scenes graded to your brand, professional edit, captions, voiceover, revisions, delivered in 24 to 48 hours.
Reels, AI UGC, product commercials and product videos start at $70. SaaS and technology content, explainers, product tours, demos, is $100, because the craft is different: screen flows, feature sequencing, UI-grade polish. Explaining software on video is its own discipline, and the tools that generate lifestyle scenes beautifully are not the tools that make a screen recording feel premium, so the pipeline work behind a SaaS video is heavier.
Volume tiers bring the per-video price down further. You keep strategy and approval; the pipeline does the production. There's no queue because the capacity scales, and there's no resignation risk because the system holds the knowledge, not a person. The way to think about the trade: you give up having "your editor" in exchange for never waiting on one. Some teams miss the relationship. Most teams, once the calendar is full for three straight months, stop missing it. The strongest case for this route isn't the per-video price at all, it's what the price does to your willingness to test. When a variant costs $70 instead of $400, you test ten hooks instead of defending one, and the testing discipline compounds into better ads every month. Price per video is a number; willingness to test is a capability.
The way to evaluate any partner, us included: ask for a sample made from your material, not a reel of their best work. Every production company's highlight reel looks great, and comparison-shopping on highlight reels is exactly how teams end up disappointed by week two of a contract. What you're actually buying is what their average Tuesday output looks like with your brand on it, and only a sample built from your product photos, your script, your captions answers that question honestly. Most partners will do one for free or cheap. If a partner won't, treat that as a data point about how confident they are in the average, not the highlight.
One more thing worth checking before you sign: who owns the underlying assets. Generated scenes, voice clones, and template files sit in murkier territory than filmed footage, and the difference between "you license the output" and "you own everything including the prompts and templates" matters if you ever switch partners.
The table that actually decides it
| Monthly need | Best route | True monthly cost |
|---|---|---|
| 1-5 videos | DIY or freelancer | $200-3,000 |
| 10-20 videos | AI production | $700-2,000 |
| 30-60 videos | AI production (volume tier) | $2,045-9,000 |
| 30-60 videos | In-house editor + tools | $4,500-9,000 |
| 30-60 videos | Freelancers | $6,000-30,000 |
At low volume the routes blur together. The gap opens at 20 videos a month and becomes a canyon at 40. Notice also what the table doesn't capture: speed and risk. AI production ships in 24 to 48 hours at every volume; every other route's turnaround degrades exactly when your volume rises, which is precisely when you can least afford it.
The hidden fees, named
Per-second pricing. Some studios quote by final duration. Reels run 30 to 60 seconds, so a "small" fee multiplies fast. Ask for per-video pricing upfront.
Revision traps. "Two free revisions, then $75 each." Revisions should be part of the job, a video that isn't right isn't finished. There's a fairness line worth holding on both sides: revisions to fix what the brief asked for are the producer's problem, revisions because the brief changed are a new job. Good partners will tell you which side of that line a request sits on before invoicing it.
Voice and music licensing. AI voices need commercial licenses. If it's not in the quote, it's on your invoice later. Same for music beds; "royalty-free" has tiered meanings, and the tier you need for paid ads is usually not the one in the free library.
Rush fees. 24-to-48-hour turnaround should be standard in 2026, not a surcharge. If a partner charges extra for speed, their pipeline is the problem.
Caption rendering and formatting. Some shops charge per platform export or per caption style. It's a small number that signals a bigger habit: monetizing the last mile instead of including it. The same habit shows up as asset fees, storage fees, and archive retrieval fees, all technically legitimate, all small, and all signs that the quoted price was a floor rather than a number.
Budgeting for 40 videos a month
Rule of thumb: 40 videos at the $70 base rate is $2,800 a month, and the marginal ladder takes that to $2,690. If every piece sat at the $150 complexity ceiling it would be $6,000; a realistic mix of reels and SaaS explainers lands between $2,690 and $4,000. Either way it is roughly one mid-level editor's salary, at two to three times the output, with no hiring risk.
The price advantage won't last forever. AI production prices will rise as quality expectations rise. Traditional prices will fall under pressure. Agencies locking in volume capacity now are buying 2027 throughput at 2026 rates. If you run an agency and want to offer video without building a production floor, our white-label AI video service for Australian agencies covers how agencies use a production partner, and the same logic applies globally.
Run your own numbers. Take your monthly video need, price it across all four routes including your own time, and compare. The difference between routes at real volume is measured in thousands of dollars a month. That comparison is worth more than this entire post, book a consultation if you want us to run it with you, or see exactly what's included per video first. For the per-video price on the two routes most readers land on, see AI product videos for Shopify brands and AI UGC ads for agencies.
The White-Label Buyer’s Checklist
Twelve questions to ask any production partner before a client’s calendar depends on them — scope, rights, confidentiality and the one answer that should end the conversation.
or download it right now — no email needed.
One email, the PDF as a link. Reply and a person answers.
The arithmetic takes an hour, and it's the highest-paid hour on your calendar this quarter. Run the numbers with us, or see exactly what's included per video first.
Drop your take 0
// no hot takes yet — start the thread 👇