QWE AI Academy The most common question people ask about a vertical series isn’t how do I make one – it’s should I make one at all, and at what budget tier? That’s what this guide answers. If you came looking for a shot-by-shot production tutorial, there are already fifty of those. This one is about the money, the compliance, and the tools – with real numbers attached.
What a vertical series actually is (the short version)
A vertical series – also called a micro-drama or vertical drama – is a serialized story shot in 9:16 for phones. Per SAG-AFTRA’s own definition, chapters run three minutes or less, end on cliffhangers, and typically start free before paid unlocks kick in around episode 5-10 (exact threshold varies by platform). A full season commonly runs 60 to 100 episodes at 60 to 120 seconds each.
$3.8 billion in 2025. Deloitte forecasts that doubles to $7.8 billion in 2026 for in-app micro-series revenue. ReelShort hit $130M in Q1 2025 alone – DramaBox right behind at $120M. The US now contributes close to half of global in-app short-drama revenue, as of 2026 reporting. This isn’t a niche format. It’s a distribution category.
The three budget tiers – pick one before you write a scene
Which tier you pick decides everything: casting options, tools, schedule, legal exposure. There’s no hybrid that dodges the tradeoffs.
| Tier | Budget | Casting | Timeline |
|---|---|---|---|
| Guild live-action | $150K-$300K | SAG-AFTRA Verticals Agreement | 8-10 shoot days + prep |
| Non-union live-action | $40K-$150K | Non-union talent only | 6-10 shoot days |
| AI-native | $20-$500/mo subs | Generated characters | Weeks per season, solo |
The trade-press median for a US vertical drama sits at $150K-$200K per production (Filmustage, 2026). That’s the live-action floor for anyone chasing a ReelShort or DramaBox deal. The AI-native tier is a different business with different failure modes – covered below.
The SAG-AFTRA Verticals Agreement, in numbers
The agreement covers projects with budgets up to $300,000, shot entirely in the US. Day rates are flat – not per chapter:
- Lead Performers: $250 for an 8-hour day, $468.75 for a 12-hour day
- Other than Lead Performers: $164 for an 8-hour day, $307.50 for a 12-hour day
P&H contributions run 20.5% on top of compensation – and 21.5% effective July 1, 2026. Old budget templates still use 20.5%. On a $200K production, that gap is real money.
The catch: Global Rule One applies to Vertical productions. A SAG-AFTRA member cannot work on a non-union vertical. So the popular “go non-union to save money” advice quietly cuts off half the trained talent pool. If your lead holds a SAG-AFTRA card and you’re non-signatory, you can’t legally cast them. No founder exception, no workaround.
Read the source directly: SAG-AFTRA Verticals Agreement Contract Bulletin (PDF). Four pages. Read them before you cast.
Worth sitting with for a moment: the format grew fast enough that a major union built a dedicated agreement for it. That’s a signal about where the money flows – and about what compliance infrastructure is now baked into the production tier.
The AI-native path – where it works, and where it doesn’t
Skip the crew, skip the location, skip the six-figure spend. Generate a season on a laptop. This is possible – but the math is more complicated than the pitch.
Runway launched Gen-4 on March 31, 2025, solving character consistency across multiple shots – a capability that had tripped up most AI video tools until then (VentureBeat covered the announcement: full write-up here). The specs matter more than the marketing, though:
Runway Gen-4 / Gen-4 Turbo:
- Clip length: 5 or 10 seconds at 24 fps
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
- Audio: not included at Gen-4 launch
Runway Gen-4.5 (expanded multi-shot, late 2025):
- Clip length: up to ~60 seconds multi-shot
- Native audio added December 11, 2025A 90-second episode on Gen-4: roughly nine stitched 10-second clips. Each references your character sheet. Each costs credits. Character drift compounds across joins – not linearly. Gen-4.5 extended clip length to around 60 seconds for multi-shot sequences, which changes the workflow. It doesn’t eliminate stitching for full episodes.
One thing that saves hours: Lock a single front-lit reference image per character before generating anything – and freeze it. Re-generating the reference midway through a season produces subtle character drift across episodes. Viewers catch it faster than you’d expect.
Community estimates put the typical iteration rate at around 3-5 generations per usable second of screen time (anecdotal; no published benchmark exists as of mid-2026). If you’re consistently under that, you’re either unusually disciplined or shipping below the platform norm. That’s an honest gap in the published docs – no platform has released a quality floor in formal spec language.
Common pitfalls the tutorials skip
Three problems quietly wreck vertical series budgets.
Shooting the whole season upfront. Revenue unlocks typically start around episode 5 to 10 (Axis AI Studios, 2026). You front $150K+ before knowing if a single viewer converts. AI-native creators can phase spend – first 10 episodes, then the rest – in a way traditional productions can’t, because there’s no location hold or cast hold forcing the full commitment. Use that.
Miscoding the P&H rate. July 1, 2026: it’s 21.5%, not 20.5%. This one lives in a bulletin, not the main rate card, which is why it gets missed.
Ignoring platform delivery specs before generating. ReelShort and DramaBox have their own delivery requirements. No universal minimum resolution is published, so check the current spec for your target platform before you lock a render pipeline – not after.
When NOT to make one
Skip verticals if any of these apply:
- Your story needs a slow burn. The format is engineered against it.
- Your hook doesn’t land in 8 seconds. The paywall math is unforgiving.
- You can’t fund episodes 1-10 at production quality. Those are the commercial product. Saving money there saves nothing.
- You’re a SAG-AFTRA member producing your own project non-union to cut costs. Global Rule One doesn’t have a founder exception.
The format rewards one specific skill set: high-frequency hook writing, disciplined asset production, retention math. It punishes almost everything else. Knowing that before you start is more useful than any production checklist.
FAQ
Is a vertical series the same thing as a TikTok series?
No. TikTok series run on open-feed ad monetization. A vertical series in the ReelShort/DramaBox sense is catalog content unlocked with paid coins. Same aspect ratio – entirely different business model.
Can I make a full 60-episode season purely with AI right now, in 2026?
Technically yes, but character consistency across 60 episodes is solvable only if you lock reference assets early and don’t deviate. The real weak points are audio, lip-sync fidelity, and physical action – all still behind live action as of mid-2026. Where AI-native works best: period drama, fantasy, heightened melodrama – genres where stylization is expected. Kitchen-sink realism exposes the gaps fastest. Solo creators shipping in those stylized genres are doing it. It’s not theoretical.
What’s the fastest way to test a concept before committing to a season?
Produce episodes 1-5. Publish to YouTube Shorts or TikTok. Watch the drop-off curve past 60 seconds. If retention holds, fund further. If it collapses in the first 15 seconds, no downstream production polish fixes that – the concept itself isn’t converting.
Next action: pick your tier – guild live-action, non-union, or AI-native – and open a spreadsheet. Line-item episodes 1-5 at that tier’s real rates using the numbers above. If the total makes you flinch, you’ve picked the wrong tier for your budget. Better to find out now.