How to Train Your Dragon (2025)

How to line your pockets

The 2025 How to Train Your Dragon film is a so-called “live action remake” of a beloved animated film (a.k.a. “LARM” if you’re a goober), but it is not like Snow White or Pinocchio for a couple of reasons. First, it is a remake of a CGI film, which is strange, as “live action remake” really means “CGI remake with human actors where applicable.” As far as I can tell, it is the first LARM of a CGI film, and — as Hunter Allen from Kinemalogue points out as an aside in his review of 2010’s Alice in Wonderland — also the first LARM of a movie initially released after the trend became profitable. The regurgitation cycle has shrunk.

Also unlike the Disney LARM dregs, How to Train Your Dragon is not painful to sit through. In fact, it is about as painless and faithful an adaptation of one of my all-time favorite animated films as I could have wished for. I don’t especially want the original changed; and director Dean DeBlois (one of the two directors of the original, along with Chris Sanders) has not changed it. This film recreates, beat for beat and almost shot for shot and line for line, the 2010 masterpiece. It honestly borders on the uncanny — almost like one of those AI recreations of classic film scenes in new styles that goes viral on film Twitter and triggers my gag reflex. This remake is less colorful than the original, and the humans are actual humans rather than 3D models, but so much of the film’s moment-to-moment timing and direction is quoted verbatim.

I’m not even entirely sure how DeBlois managed to alter the remake’s runtime so much. The sleek 98 minutes of the original is up to a nauseating 125, making me wonder if I can even trust my intuition on how similar it is. I didn’t really feel the bloating too much. The most obvious changes are a couple of added lines clarifying that the training is a competition and giving one of the teen characters a disapproving dad. I guess scenes and shots are all a little padded, and that adds up over a movie.

But I found myself curiously, even enthusiastically, soaking in familiar shots and images as I watched… and still unsatisfied. This How to Train Your Dragon is well produced and sturdily assembled, but I still felt like Edmund gobbling Turkish delight in Narnia; entirely unsated. I cannot in good conscience give the 2025 How to Train Your Dragon a thumbs up. What are we even doing here? This is a $150 million photocopy sent to theaters in exchange for a 400% ROI. It is not a reinvention or a reimagining. It is merely a faithful, reverent duplication.

The original How to Train Your Dragon is a landmark and one of the films that shaped my love of cinema when I was a young adult. Along with Pixar’s imperial era output (their first eleven films), it locked me into animation in particular. How to Train Your Dragon was a milestone not just for DreamWorks (still their creative apex in my eyes), but for CGI storytelling in general. Sanders, DeBlois, and their team — including legendary cinematographer Roger Deakins as a consultant — rendered flight and space and vast scope with unprecedented power for the medium. The film married a dangerous dragon-battling adventure with a coming-of-age arc, capturing the friction of generational change in that framework. The visual storytelling was nimble and expressive. It had wit and weight (and also some underwhelming teen sidekicks). Importantly, it had John Powell’s all-timer score, which has become a go-to reference for modern adventure composing. This remake does pretty much all of that stuff, too, I guess. At least, it goes through the motions of it.

For what it’s worth, the story still holds up. Hiccup (Mason Thames) is the gangly blacksmith’s apprentice and Viking chieftain’s son in Berk, a village constantly under siege from dragons. One day, a gadget he’s built wounds a rare and feared Night Fury, but when he finds it in the woods, helpless and injured, he does the unthinkable: he lets it go. That dragon, whom he names Toothless, soon becomes his best friend and his conduit to understanding the outside world more deeply. Through his bond with Toothless, Hiccup learns that the dragons aren’t bloodthirsty monsters but scared animals forced to feed a parasitic hive-queen dragon to survive, hence constantly scouring Berk for food. As Hiccup uses his connection with Toothless to learn more about dragons, he earns a reputation as a beast-whisperer in Viking training. When Toothless finally gets discovered, Hiccup clashes with his father, Stoick (Gerard Butler, reprising his voice role here in physical form). Meanwhile, Hiccup butts heads with other Viking teens during training, including the beautiful overachiever of the group, Astrid (Nico Parker).

The production itself is competently helmed and effective. To his credit, DeBlois, who also directed the second and third animated films in the series (and presumably will direct those sequels’ inevitable LARMs), returns with clear affection for the material rather than eyes on a paycheck. The dragon CGI, courtesy of Framestore, a British effects studio, integrates well with the humans. The gimbal-aided flight scenes don’t quite replicate the soaring magic of the original, but do as decent an imitation as you would ask for. Some of the on-location shooting in Northern Ireland and the Faroe Islands is actually more lovely than the 2010 CGI vistas. Powell returns to punch-up his great score, though it’s nearly an exact copy of his original work. (Without looking it up, I can guarantee it won’t be eligible for “Best Original Score.”) And, somehow, the added runtime really didn’t bug me too much; the story’s a but less zippy but still flows well and hits credits in under two hours.

It’s not a flawless production, though, particularly in the costuming and makeup. The Viking costumes often look like they were purchased at Spirit Halloween, and Astrid’s hair braid is distractingly bad. The distinctive color coding in the original (e.g., Hiccup in green; Astrid in light blue) has been neutered a bit in the overall brownification of the “live action” transformation.

Performance-wise, the film is a mixed bag. Mason Thames doesn’t have Jay Baruchel’s neurotic and stammering charm, but he’s a capable screen presence and generic hero. Nico Parker as Astrid is doing a theater kid performance, intense but not quite locked into the charisma and mysterious allure that the character needs. The rest of the teen ensemble — Julian Dennison, Gabriel Howell, Bronwyn James, Harry Trevaldwyn — are frankly slightly improved as human actors in the background, slightly less obnoxious and tone-rupturing than their animated counterparts. Nick Frost is having fun as Gobber, previously Craig Ferguson’s role.

Ever since Disney pioneered the CGI-driven live-action remake in the 2010s (with the Tim Burton Alice movies, then The Jungle Book), I’ve tried to keep an open mind about these LARMs. I haven’t seen it since theaters, but the 2016 Jungle Book suggested to me some possibility in interpretation — shifting stories to new tones to bring out something new in over-familiar narratives. Most do not do this, though. I don’t desire to be a kneejerk hater who prejudges these films to ignominy. As a rule of thumb, I believe it is worth engaging with these films, like any films, on their own terms. Perhaps one or two of them will justify their own existence (I hear Pete’s Dragon does). But the evidence to date does not lie: Whether they remix or remake, desecrate or worship… they all leave me cold. Maybe the big-budget LARM is not, in fact, art; maybe we’ve hit the point where it’s pure capitalism. I guess I should just stop watching them. How to Train Your Dragon might be the most frictionless specimen of the form, but it still makes me feel a little empty inside.

Is It Good?

Nearly Good (4/8)

---

Dan is the founder and head critic of The Goods. Follow Dan on Letterboxd. Join the Discord for updates and discussion.