The five films that defined 2026, and what each one proved

Five films, five different machines

2026 produced five releases large enough to define it, and the interesting thing is that no two of them worked the same way. A sanitised biopic with poor reviews, an auteur epic shot on a nearly unprojectable format, a franchise instalment at maximum scale, a horror film made for pocket change, and a legacy sequel nobody needed. All five succeeded, by entirely different mechanisms.

That is unusual. Most years are explained by one model — a run of franchise sequels, or a season where mid-budget drama returns, or a horror cycle. 2026 refuses to consolidate, which is why the coverage has been contradictory and why the year is worth examining film by film rather than as a trend. Each of the five proved something specific, and three of the five proved things the industry would rather not have learned.

The numbers, side by side

Set them side by side. Michael opened to $97.2 million from 3,955 North American theatres and crossed a billion worldwide with roughly 38% critical approval. The Odyssey cost a reported $250 million, holds around 94% approval — tied for the best of Christopher Nolan's career — and has passed $1.1 billion. Spider-Man: Brand New Day opened to $360.1 million, the largest opening weekend on record, and stands near $1.72 billion. Obsession was made for about a million dollars and took more than $240 million. The Devil Wears Prada sequel arrived twenty years late and was better than it needed to be.

How to read a year like this

Read each film against its own cost and its own audience, never against the others.

A gross figure by itself ranks these five in an order that explains nothing. Spider-Man made the most money and is the least remarkable achievement, because a franchise at that scale with that history was always going to open enormously. Obsession made the least of the four blockbusters and is by some distance the most extraordinary business outcome of the year. The useful comparison is always cost against return and expectation against result, which reorders the list completely.

Return on cost, not gross

Return on cost separates them instantly, and the ranking it produces is the opposite of the box office one.

Obsession returned something in the order of two hundred times its budget, which is not a scale that occurs in modern theatrical release. The Odyssey returned roughly four times a very large negative, which for a film of that ambition is a strong result rather than a modest one. Spider-Man returned a large multiple of a large budget in the most predictable way available. Michael's return is enormous and was purchased with a catalogue rather than with production spend. Four different businesses, described by one word.

The lesson the industry will take

The lesson the industry is most likely to take from 2026 is the wrong one.

Two of the year's biggest earners — Michael at around 38% approval and Scream 7 at roughly 30% — demonstrated that with the right title, reviews do not constrain the outcome. That is a genuine finding and it is dangerous, because it is far easier to act on than the finding that a well-made cheap film can return two hundred times its cost. Expect the sanitised, pre-sold, critic-proof project to be the format that gets greenlit, and expect the Obsession lesson to be cited constantly and acted on rarely.

What each one is like to watch

As viewing experiences the five have almost nothing in common, which is the argument for treating this as five recommendations rather than a ranking. One is a concert film with a plot, one is an endurance piece that rewards a real cinema, one is the most purely entertaining blockbuster in years, one is a tight ninety-minute genre exercise, and one is a comedy you can put on without preparing yourself for anything.

The case that none of this is a pattern

The reasonable objection is that five films is not a pattern, it is an anecdote. Any year produces a handful of large releases with different budgets and different reception, and constructing five economic models out of one summer is the kind of analysis that sounds insightful and predicts nothing about next year.

Fair, and the specific claim here is narrower than a pattern. It is that 2026 produced unusually clean natural experiments: a billion-dollar film with bad reviews, a billion-dollar film with excellent ones, a record opening built on a decade of accumulated audience, and a two-hundred-times return on a micro-budget — all in the same nine months and all visible without any interpretation. Those are data points rather than a theory, and they are worth holding precisely because they resist being combined into one.

Michael: the film that proved reviews do not matter here

Antoine Fuqua's biopic opened on 24 April to $97.2 million from 3,955 North American theatres, a record opening for the form, and has since crossed a billion dollars worldwide — passing Bohemian Rhapsody to become the highest-grossing music biopic ever made.

It did that with roughly 38% critical approval. The reviews praise Jaafar Jackson, the subject's real nephew, in his acting debut, and criticise the film around him as sanitised — specifically for omitting the child sexual abuse allegations that dominated the later part of Jackson's life.

That omission is the film's commercial strategy and its critical problem in the same decision. An estate-cooperative biopic buys the catalogue, and the catalogue is what fills the auditorium; the price is a version of the life with the difficult third of it removed. Audiences bought the music. Critics reviewed the biography.

Michael, as a thing to actually watch

It works as a performance film and struggles as a narrative one. Jaafar Jackson's physical and vocal resemblance carries long stretches, and the concert sequences are the reason the film exists — they are also where Fuqua's direction is most confident.

Between them the story moves through the expected beats with unusual smoothness, which is the sanitisation critics object to expressed as a viewing experience: nothing complicates, and so nothing accumulates.

Watch it for the staging of the music and expect nothing from the account of the man. That is a real recommendation for a specific evening and an honest description of its limits.

The Odyssey: the format became the product

Christopher Nolan's film opened on 17 July, runs 173 minutes, cost a reported $250 million — the most expensive of his career — and is the first feature shot entirely on IMAX 70mm cameras, consuming over two million feet of film.

The cast is extraordinary on paper: Matt Damon as Odysseus, Tom Holland as Telemachus, Anne Hathaway as Penelope, Robert Pattinson as Antinous, Zendaya as Athena, Charlize Theron as Calypso and Lupita Nyong'o in a dual role. It shot for 91 days across Morocco, Western Sahara, Greece, Italy, Iceland, Scotland and Malta.

It holds around 94% approval, tied for the best of Nolan's career, an A CinemaScore, and has passed $1.1 billion to become his highest-grossing film. The revealing figure is not the total but the concentration: IMAX screens, under one per cent of the total, delivered roughly a fifth of the worldwide opening.

Why The Odyssey is the year's most instructive success

Cast and source recognition explain why a billion dollars was available. They do not explain why a fifth of the opening weekend concentrated onto a handful of cinemas, or why audiences waited and travelled for a specific projection format rather than watching the same film nearby and sooner.

That behaviour — choosing where to watch, not just whether — is what every premium proposition claims to produce and almost none demonstrates. It came from a production constraint that could not be widened on demand: fewer than three dozen venues worldwide can project the format.

The lesson does not scale, which is the point. If every film shot this way, none would benefit. Scarcity was the mechanism, and scarcity is not a strategy that survives imitation.

Spider-Man: Brand New Day, the biggest and least surprising

Sony released it on 31 July. It opened to $360.1 million domestically, the largest opening weekend ever recorded, reached a billion dollars worldwide in six days, and stands at roughly $1.72 billion — $704.6 million domestic against $1.018 billion international.

Destin Daniel Cretton directed. Tom Holland returns with Zendaya, Jon Bernthal as Frank Castle and Sadie Sink as Jean Grey. The story finds Peter alone after losing his aunt and his friends, against a villain who can take control of anyone within about thirty-three feet.

The record is the least interesting thing about it. Opening weekends measure marketing and franchise position, and a fourth film in an established run was always going to open enormously. What it did with that position is the part worth discussing.

The practical stunts, and what they actually bought

Cretton brought in the Jackie Chan Stunt Team to design much of the street-level action, with Chan himself involved. Holland has said roughly 95% of the third act was performed practically. Web-swinging was rigged on wires and cranes and photographed rather than animated; pyrotechnics were detonated live, including a Glasgow opening sequence with Spider-Man swinging behind a moving truck.

None of that caused the opening. Nobody buying a ticket knew how the stunts were made, and craft cannot influence a first weekend by any mechanism. Where it shows up is later — in second weekends, in word of mouth, and in whether people return.

What it visibly bought is weight. Practically staged action has consequence because the performer's body is genuinely subject to the physics on screen, and the Jackie Chan grammar — long takes, readable geography, gags built from the environment — is legible to viewers with no interest in how it was done.

Obsession: the business story of the year

Curry Barker's horror film, distributed by Focus Features and released in May, was made for roughly a million dollars and has taken more than $240 million worldwide. That is a return in the order of two hundred times cost, which effectively does not happen in modern theatrical release.

The premise is small and executed precisely: Bear, a shy music store employee, acquires a device that makes his crush fall in love with him, with the consequences that setup promises. Inde Navarrette plays Nikki, in the performance that made her a name this year and got her cast in much larger things by August.

It is worth watching for construction rather than concept. A wish-granting premise only works if the rules are stated early and never bent, and this film holds that discipline through a third act where almost every version of this story cheats.

What Obsession did to the release calendar

The immediate consequence is visible in what studios are now buying. Low-budget horror was already the most reliable bet in the industry; a two-hundred-times return has made it the only bet several distributors want to discuss.

For viewers this is largely good news. Cheap horror is the part of the industry where first-time directors get made, and a commercially proven route to a first feature produces more interesting films within two or three years.

The risk is dilution. A form that succeeds because it is cheap and disciplined does not stay disciplined once everyone is making it, and the second wave after a phenomenon is historically much worse than the first.

The Devil Wears Prada sequel: the one nobody needed

David Frankel returned twenty years after the original, which is the kind of legacy sequel that usually exists for reasons unrelated to having something to say. This one turned out fresher and funnier than a film about the collapse of media had any right to be.

Meryl Streep gets the biggest laughs, with a new run of Miranda put-downs that land because the character was given somewhere to go rather than simply reassembled. The subject — an industry in structural decline — gives the comedy something real to push against, which is what most legacy sequels lack.

It is the most conventional film in this piece and the most reliably enjoyable. If the requirement is ninety minutes with other people and no risk of anyone hating it, this is the answer.

Why legacy sequels usually fail, and why this one did not

The standard failure is reassembly: the cast returns, the references land, and nothing has happened to anyone in the intervening years. The audience recognises the film without needing it, and the result is warm and inert.

The alternative is to let the world change and make that the subject. Twenty years of media collapse is not a backdrop here, it is the pressure the jokes work against, which gives the returning characters something to be wrong about.

That is the reproducible part. A legacy sequel works when the gap is the story rather than an inconvenience to be smoothed over, and almost none of them are built that way.

The reviews-versus-money question, stated properly

2026 gave two clean cases in each direction. Michael at 38% and Scream 7 at around 30% both posted enormous numbers. The Odyssey at 94% and Obsession, well received, also posted enormous numbers. The honest conclusion is that reviews neither guarantee nor prevent commercial success.

What reviews appear to affect is the shape of the run rather than the opening. Films with an established audience open on the title; what happens in weeks three and four is where reception shows up, and that is where a badly reviewed film with no franchise behind it has nothing to hold it.

The practical reading for a viewer is that a high gross tells you a film was wanted, not that it was good, and a low score tells you critics disliked it, not that you will.

What each film costs you in attention

The Odyssey asks the most: 173 minutes, ideally in a cinema, ideally in a format most people cannot access. It is the only one here that is materially worse at home.

Spider-Man and Prada 2 ask the least and reward a room with other people in it. Obsession is short, tight and best watched in one sitting without interruption, because its construction depends on the rules staying in the viewer's head.

Michael sits oddly: long, episodic, and one of the few films on this list that genuinely survives being watched in parts, because the concert sequences work as units.

The performances worth the entry

Jaafar Jackson in Michael is the year's most discussed debut and the reason a 38% film is defensible on any level. Whether he can act outside a role he was born adjacent to is the open question, and one the film cannot answer.

Inde Navarrette in Obsession is the performance that changed a career inside three months. A turn that survives a million-dollar production is doing something the budget did not pay for.

Meryl Streep in the Prada sequel is the least surprising and the most reliable. There is no version of this list that omits her and stays honest.

What the year says about cinemas

The single most encouraging fact of 2026 is that the two most format-dependent releases — The Odyssey in IMAX 70mm and Spider-Man with practically staged action best seen large — both performed at the top of the market. Audiences turned out for things that were worse at home.

Against that, Michael and Scream 7 demonstrated that a pre-sold title performs regardless of how it is presented, which is the argument for the opposite strategy.

Both are true simultaneously, and any exhibitor drawing a single conclusion from 2026 is reading selectively. The year rewarded spectacle and it rewarded familiarity, and it was largely indifferent to quality in between.

Ranking them honestly

By craft: The Odyssey first, Obsession second, Spider-Man third, the Prada sequel fourth, Michael fifth. By business achievement: Obsession first by an enormous margin, Michael second, The Odyssey third, Spider-Man fourth, Prada fifth.

By what most people will actually enjoy on a given evening the order changes again, and probably puts Spider-Man or the Prada sequel first. Three defensible rankings from five films is the clearest possible argument against ranked lists.

The useful question was never which is best. It is which of these five you are in the mood for, and they are different enough that the answer is usually obvious once the question is asked properly.

What 2026 changed

Three things, in descending order of confidence. Low-budget horror is now the most attractive risk-adjusted bet in the industry, and the release calendar already shows it. Format as a differentiator works when the format is genuinely scarce, and does not when it is not.

And the relationship between critical reception and commercial outcome, which was already weak, is now demonstrably absent at the top of the market for titles with a pre-existing audience.

The last of those is the one with consequences. A studio that has watched a 38% film take a billion dollars has learned something about what it needs to spend on, and quality was not on the list.

The biopic problem nobody solved

Michael sits inside a form with a structural conflict that no film has yet resolved. A music biopic needs the catalogue, the catalogue needs the estate, and the estate has an interest in the account. Every commercially viable music biopic of the last decade has made that trade, and every one has been criticised for it.

The alternative — an unauthorised film without the songs — has been attempted and does not work commercially, because the audience is arriving for the music. So the form self-selects toward the sanitised version, and criticising an individual film for it slightly misses where the constraint lives.

What makes Michael the sharpest case is the specific omission. Leaving out the allegations is not a matter of emphasis; it removes the part of the life that most of the audience arrived already knowing about, which produces the odd sensation of watching a film that is carefully avoiding something in the room.

What a $97 million opening for a biopic means

Biopics do not open at that level. The form has historically been a slow-build proposition: modest opening, long legs, awards-season attention carrying it into a second life. Michael opened like a franchise film, from 3,955 theatres.

That indicates the audience was not being persuaded by reviews or word of mouth — it was already committed before release. Jackson's catalogue is one of the very few properties with that kind of pre-sold demand, and the film's marketing correctly treated the music as the product.

The consequence is that the film was effectively review-proof by design rather than by luck. Any project able to reproduce that pre-commitment can also ignore its reception, which is a narrow category and an attractive one to studios.

Nolan's format bet, in numbers

Two million feet of film. Ninety-one shooting days across seven countries. The largest modern Viking longship reworked as an ancient Greek vessel rather than built as a digital asset. Ludwig Göransson's score built partly on bronze instruments and reconstructions of the aulos and lyre.

None of those decisions were cheap and all of them point the same direction: solve the problem in front of the camera. An IMAX 70mm camera is loud, heavy and holds only minutes of film per magazine, which forces blocking, location and design to be resolved before shooting rather than corrected afterwards.

That is why the format claim held up on screen. Audiences cannot identify a film format on sight, but they can tell when a scene was solved rather than assembled, and that legibility is what the constraint actually bought.

The Spider-Man numbers in context

A $360.1 million opening exceeds Avengers: Endgame. A billion dollars in six days is second only to a handful of releases in history. $704.6 million domestic against $1.018 billion international describes a film performing everywhere rather than in one market.

Those figures are the product of a decade of accumulated audience and a release position with almost no competition. Strip the franchise away and the same film with the same craft opens at a fraction of it.

This is worth saying plainly because the coverage has tended to attribute the record to the film's qualities. The record belongs to the franchise; the qualities belong to the film, and they are separate achievements that happened to arrive together.

Obsession and the arithmetic of risk

A studio deciding what to make compares expected return against capital at risk. On that basis a million-dollar horror film with a two-hundred-times outcome is not merely a good result; it changes the arithmetic of the entire slate, because the downside is bounded at approximately nothing.

The catch is that the outcome is not reproducible on demand. For every Obsession there are dozens of similarly budgeted horror films that return their cost and disappear, and the industry's memory selects for the exception.

What is reproducible is the risk profile rather than the result. Ten million-dollar films where one breaks out is a better business than one ten-million-dollar film, and that is the calculation studios are actually running after this year.

Comedy's quiet recovery

The Devil Wears Prada sequel belongs to a broader recovery that received almost no coverage. 2026 produced more good comedies in six months than most years manage in twelve, running from Sundance independents through Pixar to darker material from A24, Shudder and IFC.

Comedy had been the genre most damaged by the streaming transition, because it performs well at home and therefore stopped justifying a theatrical release. A year in which comedies worked in cinemas again is a structural event, not a run of luck.

The Prada sequel is the visible example because of its cast and its legacy. The recovery underneath it is the more consequential story and the one most likely to shape what gets made in 2027.

Five films, five audiences

It is worth noticing that these five barely share a viewer. The Michael audience arrived for the catalogue. The Odyssey audience arrived for a director and a format. Spider-Man's arrived for a character they have followed for a decade.

Obsession's audience arrived through word of mouth for a premise, and the Prada sequel's arrived for a comedy they already loved. Five different acquisition mechanisms, none of which transfer.

That is the strongest argument against reading 2026 as a single trend. The year did not have a hit formula; it had five separate demonstrations that different formulas still work, which is a healthier state for an industry than one dominant model.

Where the awards will land, and why it matters little

On current form The Odyssey is the only one of the five with a serious awards path, and its 94% approval and A CinemaScore make it the obvious candidate. Michael's 38% removes it from contention regardless of the debut performance at its centre.

That divergence is itself instructive. The year's most commercially successful new property and the year's most critically successful film are different films with different audiences, and neither is a corrective to the other.

Awards matter here mainly as a professional signal rather than a commercial one. None of these five needs an award to have succeeded, which is a relatively new situation and not an entirely healthy one.

What to expect in 2027

More low-budget horror, with certainty. More estate-cooperative music biopics, with near-certainty, since the form has now demonstrated a billion-dollar ceiling. More franchise continuation regardless of reception, because two films this year proved reviews do not constrain a pre-sold title.

Less certain is whether the format bet gets repeated. IMAX 70mm cannot be widened, so a second film shot that way competes for the same handful of venues, and the scarcity that made it valuable works against a follow-up.

The genuinely open question is comedy. If the recovery holds through a second year it will be treated as a return; if it does not, 2026 will be remembered as an anomaly and the genre will go back to streaming.

A closing note on how to use a list like this

Rankings of films this different are entertainment rather than guidance. A list that places a 173-minute Homeric epic above a ninety-minute horror film is expressing a preference about form, not making a judgement either of them can lose.

The version of this piece that is actually useful is the one that says what each film is for. Michael for the music. The Odyssey for a cinema. Spider-Man for a room with people in it. Obsession for a focused ninety minutes. Prada for an evening that needs nothing from you.

Read that way, five films that share no audience become five separate recommendations, which is what a year of this quality should produce.

The two films that never met the audience they deserved

Every year of this shape has casualties, and 2026's are the mid-budget films squeezed between a billion-dollar biopic and a micro-budget horror phenomenon. Neither end of that barbell leaves room for the twenty-to-sixty-million-dollar film, which is where most of the interesting work in cinema has historically been made.

The barbell is the real structural story of the year. Studios can justify almost nothing at risk or an enormous amount at risk; the middle requires a confident judgement about quality, and 2026 gave them every reason to distrust that judgement.

That is the cost hidden inside a good year. Five successful films at the extremes make the case for making more films at the extremes, and the films that would have occupied the middle simply do not get made.

What the international split tells you

Spider-Man's $1.018 billion international against $704.6 million domestic is roughly the modern norm for a franchise property. The Odyssey and Michael show different shapes, and the difference is informative.

A catalogue-driven biopic travels on the recognisability of the subject, which for Jackson is close to universal, while an auteur epic travels on the director's name and the format, which are recognised in fewer markets but very strongly where they are.

For anyone reading these numbers commercially, the split is more useful than the total. A film that earns two-thirds of its money abroad is a different business from one that earns half at home, in terms of both marketing spend and what a sequel can assume.

The one thing all five have in common

After separating them five ways, it is worth naming the single shared attribute: all five were released theatrically and behaved as theatrical films, rather than as content shot for a platform and given a token run.

That is not a small observation in 2026. The economics that produced this year — a format worth travelling for, a horror film built on word of mouth, a comedy that worked in a room, a franchise event, a concert film — all depend on people being in a building together.

Whatever else the year proved, it proved that theatrical still produces outcomes that streaming does not. Four of these five would have been ordinary at home.

What to watch, in what order

Watch them in order of how much attention you can give, not in order of how much they made.

Start with Obsession — it is short, it is the film everything else this year is measured against, and it costs you ninety minutes. Take Spider-Man next, with people in the room. Put on the Devil Wears Prada sequel on an evening that needs no preparation. Save The Odyssey for a proper cinema if one near you can project it, and for full attention if not. Leave Michael for last, treat it as a concert film, and do not expect the biography.

Revisit this list in January

Any list written in August covers two-thirds of a year, and the awards-season window has not opened. The films discussed here have been released and can be watched now, which is the argument for reading it; the ones arriving in the final quarter will change the ranking, which is the argument for not treating it as final.

The comparison worth making in twelve months is which of the five is still being discussed. Records fade quickly and craft does not, and the film with the largest opening weekend in history is not obviously the one that will be cited in five years.

Editorial conclusion

Five films, five different machines. Michael proved that an estate, a catalogue and a sanitised script beat a 38% approval rating. The Odyssey proved that a production constraint held from the first day of prep can become the thing an audience is buying. Spider-Man proved that franchise gravity at full strength still sets records, and that practical craft is visible even when it cannot be credited for the opening. Obsession proved the most valuable and least reproducible thing of all — that a disciplined million-dollar film can return two hundred times its cost. And the Prada sequel proved that a legacy sequel works when the twenty-year gap is the subject rather than an obstacle. No single lesson connects them, and pretending otherwise is how the wrong one gets learned.