The Live-Action How to Train Your Dragon (2025) Remake Succeeds Technically, But Fails Creatively
“Live action” CGI adaptations of classic animated films are an abomination. They can be successful, of course, and I’m not here to debate how they affect the film industry’s choices in general, nor how that affects artistry and culture. But I’m not here to defend them either. I was never able to survive more than a few minutes of the few I’ve attempted to watch. I couldn’t even do it for the LOLs.
But How to Train Your Dragon (2025) is a bit of a different story. Not because it’s great, but because it’s not absolutely terrible. It is far from a masterpiece. It has no reason for existing but for its cash-grabability (and who wouldn’t grab cash if they could) nostalgia baiting, and is a far cry from replicating the magic of the animated original (AO) but it’s not entirely soulless, and is enjoyable enough to watch with your kid (though, frankly, mine prefers playing with the buttons adjusting our local theater’s fancy new recliner seats).
I don’t know how you can succeed creatively with live-action remakes (LAR) of classic animated films, or even if it’s possible. Taking on some of the most beloved films of all time, you’re already up against a near-impossible standard, and there is absolutely no reason you shouldn’t be held up to it.
In the case of How to Train Your Dragon, the decision to reproduce the animated film beat by beat, without shoehorning too much unnecessary crap is a far better approach to telling the story than shoving a pile of modern-day Hollywood on top (a little doesn’t mean none). Essentially, they start off with a perfect story that has already been tested.
Keys for Switching Medium
Once you’ve made that decision, your next, and very difficult task, is to make the worldbuilding believable enough within the framework of the film to allow for immersion, so that the audience can get invested in the characters and their story.
For reproducing the modern classic, there are three elements you absolutely have to get right. First of all, most importantly, and most obviously, the dragons. The second most important thing is the kids’ ensemble and nailing their dynamics. And third, these elements must converge in perfect synergy to replicate the tone.
CGI has gotten competent enough. All you need to look at is the difference between how Game of Thrones rendered dragons in the early to mid-2010s and what House of the Dragon accomplished in 2022/2024. The tone is completely different in those cases, of course, which affects character design, but the technical capability was there. And in regard to the dragons, the LAR does an excellent job, particularly in the case of Toothless, which is, obviously, crucial, but also with the Dragon Queen, which is paramount for nailing the climactic battle.
The rest of the setting, design, costumes, and cinematography do an exceptional job of replicating the world of the animated original.
Getting the tone correct is a much more difficult job. I think that is the main reason How to Train Your Dragon doesn’t translate well to live-action. It is hard to avoid making it ridiculous*.
Starting the movie with a scene straight out of Game of Thrones’ infamously dark episode “The Long Night” (S8E3) is a terrible choice. While the film wisely avoids oversaturation, night scenes still need to be lit well enough to follow what’s happening. When everything’s too dim, it’s easy to check out. You can’t rely on nostalgia or assume the audience already remembers the context — you have to bring us into this version. It needs to stand on its own.
And once you can actually see what’s going on — since the dragons already look great — it’s time for the human stuff to kick in. That’s when the characters really need to click. They’ve got to feel grounded, and they’ve got to fit into the world the movie’s built around them.
Characters
Where the movie really falls short is the casting, acting, and directing.
Nick Frost as Gobber is absolutely fine in isolation; he just bears very little resemblance to the animated character, who is, granted, an incredibly cartoonish character. One would think Nick Frost would be up for the challenge, but no. Again, it’s not that his acting is bad by any means; it’s just the wrong choice for replicating the AO’s tone. I would have liked to have seen Brandon Gleeson or John Rays-Davies give it a go, but it might be a decade or so too late.
But here’s the real problem: the young cast is just… awful. There’s no finesse, no subtle body language, no way of speaking that gives us any sense of their inner lives. In the animated version, the voice actors made these characters feel alive — full of charm and personality. But in the live-action remake? Hiccup, Astrid, Snotlout, Fishlegs, Ruffnut, and Tuffnut are all played like cardboard cutouts. No charisma, no chemistry, nothing that even comes close to capturing the spirit of the originals.
Mason Thames as Hiccup is probably the best of the bunch — and he’s still mostly dreadful. The only thing he really sells is that he cares about Toothless, but let’s be honest: that dynamic doesn’t demand much. It’s basically horse-riding a flying dog. Jay Baruchel brought Hiccup to life with real personality and awkward charm. Thames, on the other hand, is stiff and forgettable. He completely fumbles the charisma Hiccup needs, especially when rallying his air force for the big final battle.
Jonah Hill and Christopher Mintz-Plasse (most fondly remembered as McLovin) have never really been my style when it comes to comedy, but they actually worked for their over-the-top, cartoonish characters in the (AO). Their live-action counterparts, though? Total blanks. They have nothing to do — and worse, they don’t feel like real people who belong in this world. At best, they’re a missed opportunity.
The twins, though, actually had a fun, chaotic energy in the animation — and absolutely none of that made it into the live-action version. They bring nothing to the table. I guess their only real purpose here was to check off the body-composition-swap box, so… mission accomplished, I guess? But yeah. Total dead weight.
Get Nerd Card Revoked’s stories in your inbox
Join Medium for free to get updates from this writer.
The side characters end up detracting from the main story instead of acting it out in micro or fleshing out the world.
And This is for Everything Else: Astrid
But the absolute worst of the bunch is Astrid. She’s already a tricky character — you have to handle the whole “Girlboss” thing with care, especially if the goal is for her to evolve into something deeper. She needs to come off as genuinely capable, someone who feels like a superhero: confident, charismatic, likable, and yes, attractive. She should act like she’s better than the others because she actually is, not just because the script says so. But instead, we get lazy shortcuts — a throwaway side character calling her “my girl,” or her own clunky line about being “born with nothing.” It’s all surface, no substance.
Her arc hinges on her gradual change of mind regarding Hiccup, which requires time, nuance, actual acting skills, and chemistry with the protagonist.
Barbie (2024) destroyed America Ferrera’s voice-over portrayal for me because I can’t unhear Astrid monologuing “it’s literally impossible to be a woman these days”, but she was perfectly adequate. Nico Park displays less range than the animated sheep. I would have nominated Anya Taylor-Joy when she was young enough, but, alas, I am no casting director.
And I have no idea if they’ve uglified the actress for the movie, but I love that with the look they gave her, they cast Ruth Codd** (possibly the most…interesting looking actress working today) as her mother. It explains a lot. Just not why she would be an object of infatuation (not that the boys can be very picky, or are much of a prize themselves).
And I have no idea if they intentionally “uglified” the actress for the role, but I do love that with the look they gave her, they cast Ruth Codd — one of the most unconventional-looking actresses working today — as her mother. It definitely explains a few things. Just not why Astrid is supposed to be the object of everyone’s infatuation. Then again, it’s not like the boys have high standards — or much to offer themselves.
I was going to say, at least we didn’t get Zendaya. But we kinda sorta did.
And finally, Gerard Butler is great reprising his role as Stoick in live action — his voice still works, and he looks the part. But it ends up feeling more like top-tier cosplay. Like a very committed Civil War reenactment: accurate, impressive, but still a bit hollow. The movie doesn’t know how to handle his superpowered strength either — his fights with the smaller dragons just aren’t believable, which is wild considering Hollywood pulls this off all the time in superhero movies. So no, it wasn’t inevitable.
There’s also the issue that Butler is so much better than the actor playing Hiccup. It’s like he’s acting into a void — there’s no chemistry, no spark. Without someone solid on the other end of the scene, there’s nothing for him to bounce off of. Most of their interactions land with a thud compared to the original.
When Live Action Cartoons Succeed
You can make a cartoon in live action successfully. Steven Chow has made humans into real live Loony Toons with Kung Fu Hustle (2004), RRR (2022) had people launching tigers with their arms, Paddington (2014/ 2017) delivered an astonishing emotional immersion with a coat wearing bear, Enchanted (2007) made a perfect real life Disney Princess in Amy Adams, and Netflix’s One Piece (2023) proved amine can be adapted in a way that resonates with people who can’t stand the cartoon. Then you have examples of competently generated interactions of actors with actual cartoons; it can be done in complete earnestness, like Bob Hoskins in Who Framed Roger Rabbit (1988), or completely embracing cartoonishness like Jim Carrey in Sonic (2020).
But these examples are rare. Failure is way more common.
Conclusion
The failure of How to Train Your Dragon (2025) to capture the spirit of its animated characters isn’t just about casting. It’s a fundamental misunderstanding of how and when cartoon logic, personality, and expression can translate into live action. You can’t just strip these highly expressive, stylized characters out of animation without reshaping the world and tone to fit. The result? Flat, awkward performances that miss the original’s heart and charm. When actors struggle with lines that rely on animated exaggeration, it pulls you right out of the story.
To me, this feels more like a creative stylization problem than just bad casting.
The original How to Train Your Dragon has story and character arcs that truly resonate — otherwise, why remake it? The animated film works because its characters express themselves in a world built on exaggerated physicality and emotion. The live-action remake holds together through Gerard Butler’s solid performance, the impressive recreation of Toothless, and the legacy story it follows. But sticking so closely to the original beats wasn’t just a safe move — it’s the film’s Achilles’ heel. Many emotional moments feel lost, rushed, or unearned because the faithfulness only scratches the surface. It’s less dramatic, less cathartic, less funny, and far less charming. And unlike its hero, this movie simply lacks courage.
What we end up with is a film that isn’t quite soulless but is very far from soul-ful.
*And it doesn’t help when you further shoehorn ethnic “diversity” on a small Viking cliff village, with the “backstory” that they migrated from the East. Really?! They boldly traversed the perilous seas, with dragon-infested skies, to see if they show up to WHERE THE VIKINGS LIVE, they’ll be handed visas?! And for what purpose, raising sheep?! Because, frankly, that would be an entertaining (short) film.
**Juno Usher from Mike Flanagan’s Fall of the House of Usher (2023)
