MyMuse Review 2026: the Format at Its Strangest, and Its Most Uneven

Updated August 19, 2026 · Tested by the AI Drama Reviews editorial team · 29 min read · Our score: 8.1 / 10

DramaBoxEditor's pick — biggest short-drama library, lowest price to go unlimited 9.8Excellent

• Thousands of series, roughly 200 new titles every month
• Free episodes daily, plus ad-unlocks — no card needed to start
• Unlimited plans from about $5.99 a week, the lowest entry price we found
• Runs on Android, iOS and in a desktop browser

Go to DramaBoxAffiliate link · free to download · in-app purchases available

Affiliate link. We may earn a commission if you install through it, at no extra cost to you — it never changes our scores or placements. How we test · Disclaimer

What is MyMuse?

MyMuse is a vertical short-drama app carrying an experimental, AI-produced slate, and it sits twelfth of the thirteen apps in our 2026 ranking with a score of 8.1. The format is the category standard: 9:16 video shot for a phone, episodes of roughly 60 to 120 seconds, seasons of 40 to 100 or more episodes, adding up to two or three hours of viewing.

What separates MyMuse from the apps above it is not scale, marketing or price. It is posture. DramaBox and ReelShort are commercial catalogues that happen to contain AI work; FlexTV and StoReel are AI-first businesses with a production line and a cost model to defend. MyMuse behaves more like a slate of experiments that were published rather than filtered, and it reads that way from the first browse.

MyMuse publishes no US price page, gives no catalogue count, states no refresh rate and confirms no production method. That is four missing disclosures on an app whose whole proposition is novelty, and we score it down for each of them. It is also, to be blunt, normal in this category: nine of the thirteen apps we ranked publish no official US price.

Where it fits in the wider picture is on our reviews hub, which ranks all thirteen apps side by side. MyMuse is a supplementary install, not a first one.

Is MyMuse actually AI-generated?

Yes, we classify MyMuse as an AI-native catalogue, meaning the material is generated rather than filmed. That classification is our own assessment from watching the app in August 2026 and from its positioning as an experimental AI-produced slate. It is not a percentage MyMuse has confirmed, because MyMuse confirms nothing about how its titles are made.

Be clear about the limit of that statement. No authoritative dataset exists showing what share of any app's catalogue is AI-generated, on MyMuse or anywhere else. Anyone who quotes you a figure for this app invented it. What we can say is what the screen shows, and on MyMuse the screen shows generated video with the artefact signature that generated video has.

The industrial context explains why an app like this exists at all. The China Netcasting Services Association reported that more than 95% of new Chinese microdramas in the first quarter of 2026 were AI-generated, and DataEye counted roughly 221,900 AI titles published on Douyin in the first half of 2026, about 1,200 a day. Of those, around 1,055, or 0.47%, passed 100 million views. Production is now cheap enough that publishing an experiment costs almost nothing, and MyMuse is what a catalogue looks like when that is true.

If you want the mechanics rather than the market, we set out the full production chain in how AI dramas are made.

Why quality swings so hard between titles

MyMuse quality varies more between its own titles than the average quality of most rival apps varies between each other, and the cause is that almost everything determining the result is decided before a single frame is generated. The model is the least interesting variable. Genre choice, character reference discipline, shot length and the number of human hours spent on continuity decide whether an episode holds together.

This is the section of this review that exists nowhere else on the site, because MyMuse is the app where the mechanism is visible without instrumentation. Watch four titles in an evening and you will see the full range of what current generative production can and cannot do, which is genuinely useful if that is what you came for and infuriating if it is not.

Read that pipeline again with variance in mind. Stage 4, the generation step, is the one everybody pictures and the one that differs least between a good MyMuse title and a bad one, because everyone is drawing on a similar generation of models. The quality difference is manufactured at stages 1, 2, 3 and 6. A locked character bible at stage 3 and disciplined continuity editing at stage 6 are what separate the episodes that work from the ones that do not.

Driver one: genre fit

Stylised fantasy hides generative artefacts; contemporary human drama exposes them. A viewer watching an invented world has no reference for what a robe, a spirit beast or a floating palace is supposed to do, so an inconsistency reads as style. A viewer watching an argument in a modern kitchen knows precisely how a hand grips a mug, how a shirt collar sits and how a face stays the same face across a cut.

That is why the strongest MyMuse titles we watched were fantasy and the weakest were contemporary office and family drama, and it matches the only reviewer account we have. Jenny Cooper of Vertical Drama Love reported on June 24, 2026 that she watched 25 AI titles, finished three, concluded that story remained decisive, and named fantasy the strongest category. One reviewer's experience is not a dataset, and we cite it as testimony rather than measurement, but the mechanism behind it is not mysterious.

Genre by genre, we set out where generative video is comfortable and where it is not in the AI drama genre guide.

Driver two: how much human continuity work the title received

Human continuity editing is the most expensive input in an AI drama and the first thing cut when a title is a test rather than a bet. Repairing artefacts, fixing lip-sync, smoothing pacing and rejecting shots that do not match is labour, and labour is what the technology was adopted to remove. FlexTV vice-president Tang Tang told MIT Technology Review on May 15, 2026 that moving North American production to AI cut costs by 80 to 90% and schedules from three or four months to under one, with teams of about ten people. That is a vendor describing its own results, and it is also a description of how little human time each title now gets.

On a catalogue like MyMuse, that budget is not spread evenly. Some titles clearly received shot-level review and some clearly did not, and you can tell which within thirty seconds. The metric that matters inside a production is cost per accepted shot rather than cost per generation, because a 90-second episode can require many minutes of raw generated material to fill.

Driver three: shot length discipline

Shots that stay under about 10 seconds survive; shots that push past it drift. This is a hard model constraint rather than an aesthetic preference. ByteDance's Jimeng generates roughly 12 seconds at a time, or 30 seconds with motion templates; Kling reaches around three minutes; Sora 2 and Runway Gen-4.5 top out at 1080p. A scene is assembled from multiple short generations, never one long render.

A disciplined MyMuse title cuts before the model loses the thread. An undisciplined one holds a shot for fifteen seconds and lets you watch a face reorganise itself in real time. Nothing about the underlying technology changed between those two episodes. Somebody simply made a different editing decision, and the format punished it.

Driver four: the character bible and reference images

A locked character bible with fixed reference images is the single highest-leverage decision in the whole chain, and it is invisible to the viewer until it is missing. Character reference images, fixed seeds, first-and-last-frame controls and a locked bible reduce identity drift substantially. They do not eliminate it. Google's Veo 3.1 exposes reference images, first and last frames, camera controls and scene extension precisely because this is the problem the whole field is trying to solve.

There is a second-order effect worth naming, because it shapes how the whole MyMuse catalogue feels. Models trained on similar data produce similar faces, so different series can feel cast from the same handful of people. On an app with an unusually experimental slate that sameness is more noticeable, not less, because the stories are more varied than the faces telling them. We explain the underlying mechanism in why AI drama characters change between shots.

Now put the two lists together, and MyMuse stops being random. A fantasy episode with a locked bible, short shots and one continuity pass will show you perhaps two of those eight modes and hide both. A contemporary drama generated quickly with no bible, long takes and no repair pass will show you six of them in ninety seconds. Same app, same week, same technology.

None of this is a MyMuse secret. It is the standard mitigation set that the research literature lists and that every producer in the category knows: character reference images, a locked character bible, fixed seeds, first-and-last-frame controls, shorter shots, shot-level human review and continuity editing. These reduce the failure modes and do not remove them. What makes MyMuse worth writing about is that it applies them inconsistently enough for you to see the difference they make.

How does MyMuse score, axis by axis?

MyMuse scores 8.1 out of 10, held down by billing transparency at 5.5 and by a catalogue that is neither large nor consistently refreshed. Weights and protocol are on our method page.

One note on the production quality score, because a single number is a poor instrument for this app. A 7.4 here does not describe a consistently mediocre catalogue. It describes a catalogue where some titles would score above 8.5 and others below 5, averaged into a figure that describes no individual episode you will actually watch. We say so rather than pretending the mean is the experience.

How much does MyMuse cost to finish one season?

MyMuse publishes no official US price, and this review therefore states none. We found no pricing page, no committed coin table and no store figure we were prepared to put in print. Under our own rules that absence is not a gap to be filled with an estimate, it is a finding, and it is why billing transparency scores 5.5.

What we can give you is the category context, clearly labelled as category context rather than as MyMuse pricing. Coins across the apps we tested run about $0.20 to $0.50 an episode, which puts a 70 to 80 episode season somewhere between $14 and $40. Weekly unlimited plans across the category span $5.99 to $29.99. MyMuse operates a coins-plus-subscription model, so both of those shapes are relevant, but neither is a quote.

Verify at checkout is not boilerplate on this page. It is the entire pricing advice, because the number the app shows you is the only number that exists.

The arithmetic of coins against weekly plans, worked through for every app in the ranking, is in what AI dramas actually cost.

How much can you watch on MyMuse for free?

MyMuse offers a limited free layer of opening episodes, and on this app the free tier is not a trial, it is the product's quality control. Everywhere else in this category we tell you to use the free episodes to decide whether you like the format. Here you use them to decide whether you like the title, because the answer changes from title to title.

The structure is the category standard. The first few episodes are open, the wall arrives at the moment a cliffhanger has made you least rational, and unlocking continues by coins or subscription. That placement is deliberate and it is not a MyMuse invention; every app in this ranking does it.

The practical routine we would use, and did: open four titles, watch the free episodes of each, and only then decide whether any of them justifies a payment. If two of the four are still holding your attention at the wall, MyMuse is worth a week. If none is, you have learned something valuable for free, which is more than most paid experiments deliver.

What is actually in the MyMuse catalogue?

MyMuse runs a small, experimental catalogue with no published title count and no stated refresh rate, and it scores 7.0 on that axis for exactly those reasons. Against DramaBox, which adds roughly 200 new titles a month across thousands of series, MyMuse is a boutique. Against StoReel, which raised $34 million and targets 100 AI dramas a month according to 36Kr on March 26, 2026, it is also a boutique.

The compensation is that a MyMuse browse does not feel like every other browse. The majors converge on the same shelves because those shelves earn: billionaire romance, revenge, werewolf. MyMuse attempts things the commercial catalogues would not fund, and about half of them do not work.

One structural fact worth repeating on every review we write: catalogues in this category never overlap. Every platform finances its own titles, none licenses to another, and there is no cross-app search. Whatever is on MyMuse exists only on MyMuse, which is the only real argument for keeping it installed alongside a bigger app.

How good is the video, audio and dubbing on MyMuse?

At its best MyMuse looks better than its ranking position suggests, and at its worst it is the roughest video in this ranking. The best titles are clean 1080p-class output with composed framing, consistent characters and cutting that respects the model's limits. The worst show three or four of the eight failure modes above inside the first minute.

Audio follows the same pattern and is the more reliable tell. Where a title has had a voice pass recorded and corrected against picture, dialogue lands. Where it has not, lip movement drifts a beat behind the line and emphasis falls in the wrong place, which is the same complaint viewers make about heavily dubbed live-action catalogues for a different underlying reason.

What to look at in the first thirty seconds

Check three things and you can predict the rest of the episode. First, hands: if a character is handling an object in a centred close-up and it holds together, someone cared. Second, cut rhythm: shots consistently ending before about ten seconds means the shot list was written for the technology. Third, background signage: unreadable pseudo-letters in a modern setting means nobody did a repair pass, and nobody doing a repair pass means the continuity errors later in the episode were not caught either.

That thirty-second test is worth more on MyMuse than any review score, including ours. We can tell you the catalogue is variable. Only the episode in front of you can tell you which half of it you are in.

Where does MyMuse play?

MyMuse is a phone app: Android and iOS, and effectively nothing else. That is enough for the format, since 9:16 vertical video is designed for exactly one screen shape, but it is less coverage than the apps above it in the ranking offer and it costs MyMuse two points on that axis.

How does MyMuse compare with the alternatives?

The three apps to weigh MyMuse against are MoboDrama, FlexTV and StoReel, because they are the other places a US viewer can watch a substantially AI-produced slate. On score, MyMuse is last of the four. On the specific job of showing you what generative drama currently is, it is not.

If you install exactly one AI-first app, install FlexTV. It is the platform whose move to AI production is documented rather than asserted, its vice-president Tang Tang having described the 80 to 90% cost reduction and sub-one-month schedules to MIT Technology Review on May 15, 2026. It is also more consistent than MyMuse, which for most viewers matters more than range.

If you want to make something rather than only watch it, install StoReel. It is the only app here with a creation tool, it publishes an actual price, and one reported case put four one-minute episodes at $150. It is also among the most expensive weekly plans in the category, which is the trade.

MoboDrama is the closest comparison and the hardest call. It scores 8.2 against MyMuse's 8.1, a difference small enough that we would not argue with anyone who reversed them. MoboDrama claims a fully AI catalogue and runs at the smallest scale we tested; MyMuse attempts more and therefore misses more. Try both free tiers in one evening and let the titles decide.

What does MyMuse do well?

What does MyMuse do badly?

Three of those weaknesses deserve more than a bullet, because they are the ones that cost you either money or an evening.

The price opacity is worse here than on a bigger app. When DramaBox declines to publish a US price, the damage is limited by the fact that observed figures cluster around $5.99 a week and the catalogue justifies it either way. When MyMuse declines, you are being asked to enter payment details for a catalogue whose quality you cannot predict, at a price you cannot check, for a season length nobody has stated. That is three unknowns stacked, and we score it as such.

The variance is not evenly distributed across the things you care about. If the weak titles were weak at random, sampling would be cheap. They are not: the weakness concentrates in contemporary settings, dialogue-heavy scenes and anything involving hands and props, which is to say in most of what makes drama emotionally legible. The strong titles concentrate in fantasy. If fantasy is not your genre, MyMuse has much less to offer you than its score implies.

Nothing on the app helps you tell the two apart. There is no quality signal, no editorial curation you can see, no badge, no ordering that puts the strong work first. The thirty-second test earlier in this review exists because the app supplies no alternative. On a catalogue this variable, the absence of curation is a product failure and not a neutral choice.

For completeness, the failure modes described in this review are not unique to MyMuse and we would say the same of any AI-native slate. What is specific to MyMuse is the width of the gap between its best and worst work, and the fact that it publishes nothing that would help you land on the right side of it.

Who should install MyMuse, and who should not?

Install it if…

• You are curious about AI-produced drama as a thing in itself and want to see the full range rather than a curated best-of.
• Fantasy, mythological or costume drama is your genre. That is where this app is genuinely good.
• You already have a main app installed and want something different for the evenings the majors bore you.
• You are researching the format, writing about it, or making it yourself and need to see what the current tooling produces without a studio in the way.
• You are comfortable abandoning a title after ninety seconds without feeling you wasted anything.

Skip it if…

• You want a dependable evening of television. The variance will find you.
• You mainly watch contemporary romance, office or family drama. That is the app's weakest ground.
• You want to know what you will pay before you install. MyMuse does not publish it, and we will not guess.
• You watch on a TV, a tablet or a laptop. This is a phone app.
• You want the cheapest route to a lot of viewing. That is DramaBox, at about $5.99 for a week that finishes a whole season.

How do you try MyMuse without paying?

1. Install and stay on the free layer first. No card is needed to watch the opening episodes, and on this app those episodes are doing real work for you.
2. Turn off in-app purchases before you browse if the phone is shared, or if you know your own resolve at 1am on a cliffhanger.
3. Open four titles, not one. Start with fantasy, then take one contemporary title deliberately so you see the range for yourself.
4. Run the thirty-second test on each. Hands in close-up, cut rhythm under about ten seconds, background signage. Two failures out of three and you can close it without guilt.
5. Only then look at the price the app quotes you. It is the only price that exists, so read it carefully and screenshot it.
6. If you pay, buy time rather than coins where you have the choice, and cancel in the store on the same day so the period you paid for still runs out on schedule.

New to the format entirely? Start with how to start watching AI dramas instead, and come to MyMuse second.

How do you cancel MyMuse?

Cancel in the store, not inside the app. This is the most common billing mistake in the category and it is why people are charged for a second period they did not want.

1. iPhone and iPad: Settings, then your name at the top, then Subscriptions, then MyMuse, then Cancel Subscription.
2. Android: Play Store, then your profile icon, then Payments and subscriptions, then Subscriptions, then MyMuse, then Cancel subscription.
3. Confirm it worked. The store should show an expiry date rather than a renewal date. If it still says it renews, it will renew.
4. Keep the receipt. With no published price page, your store receipt is the only record of what you were actually charged, and you will need it if you ever dispute the amount.

You keep access until the period you already paid for ends, so cancelling immediately costs you nothing. The full walkthrough for every app is in how to cancel a short drama subscription.

Frequently asked questions

Is MyMuse worth installing? The verdict

MyMuse scores 8.1 and we recommend it to a narrow audience with an unusual reason: it is the clearest window in this ranking onto what AI-produced drama actually is right now, unsmoothed by a commercial editor. The best titles on it are better than the ranking position implies. The worst are the roughest video we watched anywhere. Both facts come from the same catalogue in the same week.

The variance is the finding, not the complaint. Quality in this format is decided by genre fit, a locked character bible, shot length discipline and the number of human hours spent on continuity, and MyMuse applies those unevenly enough that you can watch the difference they make in a single evening. Nobody else's catalogue teaches that as efficiently, which is worth something even though it is not worth a first install.

Its worst failing is that it will not tell you what it charges. No US price page, no coin table, no figure we were willing to print. We score billing transparency at 5.5 and we would rather leave a hole in this review than fill it with an estimate, because an estimate is exactly the thing that would cost you money if it were wrong.

Install it second, not first. Have a main app for the evenings you want to be entertained reliably, and keep MyMuse for the evenings you want to be surprised. Exhaust the free tier, run the thirty-second test, and treat any subscription as a one-week experiment with a cancellation already booked.

DramaBoxEditor's pick — biggest short-drama library, lowest price to go unlimited 9.8Excellent

• Thousands of series, roughly 200 new titles every month
• Free episodes daily, plus ad-unlocks — no card needed to start
• Unlimited plans from about $5.99 a week, the lowest entry price we found
• Runs on Android, iOS and in a desktop browser

Go to DramaBoxAffiliate link · free to download · in-app purchases available

Affiliate link. We may earn a commission if you install through it, at no extra cost to you — it never changes our scores or placements. How we test · Disclaimer

Related reading

Sources

• MyMuse — app store listing and in-app browsing on a US account, August 19, 2026. MyMuse publishes no official US pricing page, no coin table and no catalogue count. The AI-native classification on this page is our own assessment from viewing, not a company statement.
• Our own testing — installation on Android and iOS US accounts, free-tier exhaustion, title-by-title viewing across genres and cancellation checks, August 2026. Protocol on how we test.
• Vertical Drama Love, reviewer Jenny Cooper (June 24, 2026) — 25 AI titles watched, three finished, story described as decisive and fantasy as the strongest category. Cited as one reviewer's stated experience, not as survey data.
• MIT Technology Review (May 15, 2026) — FlexTV vice-president Tang Tang on the move of North American production to AI: costs down 80 to 90%, schedules from three or four months to under one, teams of about ten. A vendor describing its own results.
• China Netcasting Services Association via Global Times (2026) — more than 95% of new Chinese microdramas in Q1 2026 were AI-generated.
• DataEye (2026) — roughly 221,900 AI dramas published on Douyin in the first half of 2026, about 1,200 a day, of which around 1,055 (0.47%) passed 100 million views.
• ACM, A Survey: Spatiotemporal Consistency in Video Generation (May 18, 2026), and OpenReview on narrative and temporal consistency in long video (March 26, 2026) — the documented failure modes and the mitigations that reduce them.
• arXiv:2510.04999, Bridging Text and Video Generation: A Survey (October 6, 2025) — model capabilities and limits underlying the shot-length constraint.
• Model documentation and reporting on generation limits — ByteDance Jimeng at roughly 12 seconds per generation and 30 with motion templates, Kling to about three minutes, Sora 2 and Runway Gen-4.5 to 1080p, and Google Veo 3.1 reference images, first and last frame controls and scene extension.
• 36Kr (March 26, 2026) — StoReel raised $34 million with a target of 100 AI dramas a month, used here only as the comparison point in the alternatives table.
• PetaPixel and Today (August 2026) — director Qingge Gao on microdrama viewers not caring whether a title is AI-made. Cited as expert opinion, not measurement.

Affiliate disclosure, plainly stated. We earn nothing from MyMuse. There is no affiliate programme behind any MyMuse link on this page and no commercial relationship of any kind between us and this app. The only app on this site that pays us is DramaBox, whose promotional block appears above, and that arrangement did not stop us ranking DramaBox first on price while calling its refusal to publish a US price its worst failing. If commission drove our verdicts, an app that pays us nothing would not be reviewed at this length. Accuracy notice. Prices, coin values, free-episode allowances and catalogues change without notice and vary by region, device, store and promotion. Everything here was verified on August 19, 2026, and MyMuse published no US price on that date, which is why this review quotes none. The figure the app shows you at checkout is the only binding one. Independence. We are not affiliated with MyMuse and this page is not an official support channel. For account, refund or billing problems, contact MyMuse through the app store you purchased from.