James Gunn’s Superman (2025)

I Still Believe

James Gunn understands what Richard Donner understood, which is an uncommon thing in this genre. He believes that earnestness in a cynical age is the actually radical posture, and he’s made the first Superman film since 1978 that operates in that belief. The film has real virtues. It also has serious problems, almost all of which cluster around the same structural failure. I want to argue both halves, because the film deserves serious consideration.

Begin with the casting, because it is the first place the film earns its standing. David Corenswet is the first actor since Christopher Reeve who is actually Superman to me. Routh did a Reeve impression, which was respectful but airless. Cavill was asked to play Superman as a brooding demigod in a universe that distrusted him, and he committed to that material, but the material itself was working against the character’s essence. Corenswet has a quality of unguardedness the intervening Supermen weren’t allowed. He is the first actor since Reeve permitted to be earnest without irony, and physically imposing without menace. That combination is harder than it appears. Many actors can manage one of the two. Corenswet manages both, and he does it with a contemporary register — less formal, less theatrical, and more continuous between Clark and Superman — that speaks to a generation that reads the old Clark/Superman split as slightly camp. Whether that contemporary register is a gain or a loss is a real question the film doesn’t entirely resolve, but Corenswet’s performance isn’t the place the film falls short.

Rachel Brosnahan’s Lois is a different inheritance. She’s working in the post-Kidder era, knowing she can’t be Margot Kidder and refusing to try. What she does instead is honor what Kidder built without imitating it — same values, same professional seriousness, and the same refusal to be managed, but expressed through a different person who moves differently and speaks differently. She plays Lois as a working journalist first and a love interest second, which is closer to Kidder than people give her credit for, and she does it without the neuroses, which is a defensible reading of how the character might land in 2025. I’m, as a matter of personal disposition, in for anything Brosnahan does. But the performance does not need that bias to work. It works on its own terms.

There’s a scene in this film I’ll defend forever, and it’s the reason I’ll keep affection for the picture no matter what else I say. Clark and Lois are in his apartment after a fight. He’s been out looking for the dog. They’re talking about punk rock, of all things, and Lois is articulating the thing she said in the fight that wounded him — that they are too different, that he sees the best in everyone he looks at, that she’s the skeptic and the questioner and that the gap between those orientations isn’t closeable. He is sitting there in the costume of a god. She’s in her own clothes, and yet she’s the one with armor on. He has the powers but not the defenses. And when he tells her — gently, almost as a private observation — that maybe seeing people that way is the real punk rock, he’s offering her a reframe of her own identity. The thing she thought made her tough, her suspicion and her edge, he is suggesting might be the conformist position now. And he’s right. In 2026, sincerity is the countercultural move. Cynicism is the house style.

That scene is the heart of the film and the moment Gunn most fully trusts his material. It’s also the film telling you what it is. Gunn is making a punk rock Superman movie — not punk in aesthetic, but punk in the sense of refusing the dominant mode. The dominant mode for two decades has been the deconstructed superhero, the morally compromised hero, the grimdark inversion. Gunn’s move is to play it straight, and the kitchen scene with Lois is where he says so out loud. It’s the best argument the film makes for itself.

And the dog. I have to talk about the dog. I’ll be mocked for this, and I don’t care. Krypto is the most convincing CGI creature I’ve seen in years, and the reason is that the animators gave him actual canine physics. He tugs against Superman’s grip with real torque. His weight shifts believably when he lands. He doesn’t move with the floaty over-articulation that gives most CGI animals away. He moves like a dog who doesn’t know he has powers, which is exactly right. The cape would be a disaster on a lesser production. Here it reads because the dog underneath the cape was solved first.

But the dog is doing more than technical work. He’s doing thematic work. Superman with Krypto is a man who can’t fully control his own household. The most powerful being on Earth is being dragged through his apartment by an animal who won’t listen. That isn’t a gag, it’s a statement. Power doesn’t insulate this man from the small frictions of being alive. He still has to chase his dog. He still has to put his boots on, as the film’s central speech puts it. The dog grounds him in domesticity and in a life that has both texture and resistance. Without the dog, the speech about getting up and doing the work anyway is abstract. With the dog, you’ve already seen the man it describes. That’s real screenwriting, and it’s one of the things the film gets exactly right.

Now the harder half. The film’s failures aren’t random. They cluster around a single structural problem, and once you see the pattern, you see it everywhere.

The problem is that Gunn was given the architect’s job. He isn’t just making a Superman movie; he’s building the entry point for a whole DC universe, and that obligation has taxed almost every storytelling decision in the picture. Rick Flag Sr. exists for franchise reasons. He has no actual relationship to Superman, no dramatic function in this film, and every time he enters the frame, the picture goes inert. He’s connective tissue for movies that haven’t been made yet. The Justice Gang — Mister Terrific, Hawkgirl, Green Lantern — are the same problem repeatedly. They don’t test Superman, they don’t change him, they don’t reveal him. They show up, they fight, they leave. They could be removed entirely, and Superman’s journey would be identical. That’s the screenwriting test, and they fail it. The 1978 film passes that test for every speaking part. Lois changes him. Luthor opposes him. Jor-El instructs him. The Kents formed him. Even Otis and Eve serve Luthor’s characterization. Nobody is there for nobody. Gunn’s film has fifty speaking parts, and most of them are functions.

The cost is most painful in Lex Luthor. Nicholas Hoult is a real actor with miles of range — The Great is sufficient evidence on its own — and he’s doing exactly what he was directed to do. The miscasting isn’t in his instincts. It’s that a high-range actor was handed a low-range part. Hackman’s Luthor was, before anything else, fun to be around. He was vain, theatrical, snide, weirdly domestic with Otis and Eve, simultaneously a megalomaniac and a man complaining about the wallpaper. Donner gave Hackman leisure — scenes that existed simply to let us laugh at Luthor’s personality. Hoult never gets that leisure. His Luthor is a function, the engine that produces the kryptonite, the clone, the pocket universe, and the smear campaign. He has subordinates but no Otis. He has rage but no leisure. And leisure is where Hackman built the character. The cardboard quality isn’t Hoult’s failure — it’s what the script gave him room for. Luthor’s screen time was taxed by the franchise, and the result is a Honda Civic performance from a Formula One driver.

This is the same disease that makes the contemporary Eve Teschmacher, and here the comparison is most unforgiving. Valerie Perrine’s Eve was the moral measuring stick of the 1978 film. Her presence next to Luthor told you that even inside his operation there was a person who knew what was right and was choosing wrong, until the moment when the cost got real and she chose right. The 2025 version is an influencer. She’s a satirical type rather than a person, vain and witless and played for the camera’s contempt, and there isn’t a scene where her humanity surfaces. The film’s posture toward her is the same camera-above posture it sometimes takes toward Ma Kent during the phone scene — and it’s the posture I most want Gunn to abandon. Hackman’s Luthor was given complexity by the people around him. Hoult’s Luthor is surrounded by assets on a balance sheet. So Luthor himself becomes thinner, because his ecosystem is thinner.

The Kents are the film’s tonal failure in microcosm, and the failure is not where critics usually locate it. It isn’t that Gunn made them fallible or rural or unsophisticated — that’s defensible and even necessary for the thesis. The failure is that the film can’t decide whether to love them or use them as comic relief, and it tries to do both in adjacent scenes without modulating the register. The boot line is perfect. Ma Kent telling her son she will go get his boots, knowing what he is about to walk into, doing the one thing she can do, expressing love through the only vocabulary available to her — that is a great moment. The porch scene with Pa Kent is a great moment. His speech about choices, about parents giving children the tools to go out and become who they are, about the fact that what makes you who you are is what you do — that is the film’s actual answer to the nature-versus-nurture question. These scenes treat the Kents as people whose limitations are inseparable from their wisdom.

Earlier, the phone scene happens. Here, the film betrays them. Ma Kent is not a woman whose love exceeds her vocabulary; she’s a rube being laughed at by the camera. The framing changes. We’re no longer with her; we are above her. The problem isn’t that the scene is unrealistic — plenty of people are bad with technology — the problem is that the film’s posture toward her has shifted, and shifted in a way that contradicts everything the porch scene established. You can’t ask the audience to take the Kents’ wisdom seriously and then invite them to laugh at the Kents’ stupidity because wisdom and stupidity aren’t separable that way. They’re the same person. This is a Gunn problem more broadly. It’s the same instinct that gave us the Guardians films, where genuine emotional beats sit next to needle-drop irony, and most of the time he gets away with it because those films operate in a world that licenses the whiplash. Superman doesn’t license it. Superman, as Donner understood, requires a unified tonal field. The moment you laugh at anyone in the Kent family, you’ve broken something structural about who Clark is and where he comes from.

A major action sequence is the film’s largest single failure. The pocket universe sequence has to function as the resolution of the kryptonite-imprisonment arc and as the bridge to the third act, and it carries neither weight. The visual design is muddy because the effects team is asked to render too many concepts in the same frame: dimensional rift, ion river, black hole, baby in peril, Superman straining against physics. There is no through-line for the eye. Compare it to Donner’s climax, which is essentially Superman flying around the Earth — one image, one action, total visual clarity, and you feel every second of it. Gunn’s sequence has more events and less momentum, which is the signature problem of the contemporary blockbuster climax.

And inside that sequence is one of the sharpest images in any superhero film of the last decade — the troll farm of monkeys at keyboards, generating bot content and chatbot replies. That is a brilliant satirical idea, and it deserved to be the spine of Luthor’s operation, the running visual joke that built up the film and paid off in the third act. Instead, it is a sight gag the film walks away from. I laughed. The people I was with did not even register it. That tells you something. When a genuinely funny image doesn’t land in the room, it’s almost always because the film hasn’t done the work to set it up. The joke was too good for its placement.

The desert battle has the same problem. It looks like a backlot, and the reason is almost certainly that the budget got allocated across too many setpieces. When you have a pocket universe sequence, a Metropolis attack, a kaiju fight, and an international desert battle, the VFX budget per shot drops, the location work gets compromised, and the visual field flattens. Compare to Donner’s helicopter rescue — one sequence the film commits to fully, real Manhattan, real heights, real wind, a single clear physical event. The whole 1978 film is built so that sequence can land. Gunn’s film is built so eight sequences can occur, and none of them lands fully.

The Lois-Clark punk rock scene is the perfect crystallization of this problem, and it is a bug masquerading as a feature. Gunn presents it as visual wit — the world-saving happening offscreen while the real movie happens in the kitchen — and on paper that is a beautiful idea. In execution, the background visuals are loud enough to fight the foreground for attention but vague enough that nobody can parse what is happening. The audience splits. Some viewers try to read the imps in the window and miss the kitchen. Others stay with the kitchen and find the window distracting. Almost nobody does what Gunn wants, which is read both layers as a single ironic composition. The scene would be stronger with the window dark. The intimacy would be total.

The diagnosis is this: Gunn’s instincts are right, and his execution betrays his instincts because the franchise architecture won’t let him commit to enough. He believes in earnestness but stages it with ironic backgrounds. He believes in Superman but gives him a team. He believes in Luthor but splits him with Flag. He believes in the kitchen scene but puts a battle in the window. The film can’t stay still long enough to be itself.

This is the deepest argument I want to make about contemporary blockbuster cinema, and Superman is the cleanest example of it I have seen in years. A film used to be a closed object — a beginning, a middle, an end, a complete experience. Now it’s a node. It has obligations to the films before it and the films after it, and those obligations are paid in screen time that used to belong to the story. The audience feels this even when they can’t articulate it. They feel the film’s attention drifting away from them and toward an invisible spreadsheet. Gunn knows better than this. The Suicide Squad worked because it was allowed to be its own thing. Peacemaker worked because it was a contained character study. The moment he became the architect of the whole DCU, his films inherited the architectural problem. Superman is the first film where you can see the costs.

And yet, I’m ending where I began. Corenswet is Superman. Brosnahan is Lois. The kitchen scene is the kitchen scene. The dog is the dog. The boot line is the boot line. There are minutes in this film where the right values, the right register, the right faith in the audience all align, and in those minutes you feel the thing the 1978 film made you feel. Gunn’s the first director in two decades who has even attempted that frequency, and he reaches it more often than his contemporaries do. The film isn’t even half of what Donner made. But it is the only contemporary Superman that has remembered what Superman is for, and that a lot. The reach exceeds the grasp, but the reach itself is honorable. I’ll take that over polished cynicism any day of the week.

Half of what Donner made is still more than almost anyone else is making. I will be back for the next one.