Released in 2017 and developed by Vicarious Visions and published by Activision, Crash Bandicoot: N. Sane Trilogy is a remastered compilation of Crash Bandicoot, Crash Bandicoot 2: Cortex Strikes Back, and Crash Bandicoot 3: Warped. It launched first on PS4 and later expanded to major modern platforms including Nintendo Switch, Xbox, and PC, which helped it become a shared “hub” for multiple Crash speedrun scenes under one umbrella release.
Speedrunning the trilogy works because the games are built from short, discrete stages with clear endpoints, fast restarts, and highly repeatable platforming challenges. Runs reward tight movement and clean lines through fixed-camera obstacle courses, plus smart risk management around cycle-heavy hazards, crates, and masks. Because each of the three games has its own feel, the trilogy naturally supports both “single game” mastery and marathon style routes that test consistency across different movement demands and level gimmicks.
What makes N. Sane Trilogy distinct from the original PlayStation-era runs is that it is not just the same games with prettier visuals. The remake’s physics and feel are its own ecosystem, including jump behavior that has been widely discussed as meaningfully different from the originals, which matters a lot when speedroutes are built around landing precision and momentum control.
At the scene level, the trilogy also has a clear “public record” structure, with boards organized by each individual game and additional combined categories. Timing culture is shaped by the practical reality that loads vary across hardware, so runs are commonly tracked with loadless timing alongside an “RTA with loads,” keeping comparisons meaningful across platforms and capture setups.
When Crash Bandicoot: N. Sane Trilogy arrived as a rebuilt compilation by Vicarious Visions and Activision, its speedrunning scene formed fast because runners already understood the bones of Crash Bandicoot, Crash Bandicoot 2: Cortex Strikes Back, and Crash Bandicoot 3: Warped. Early route work was less about learning the games from scratch and more about answering a new question: how do those familiar levels behave inside a modern remake with different movement feel, new loading behavior, and a new set of optimal risks?
The earliest public “record book” for the trilogy consolidated around Speedrun.com, which organized the scene into distinct leaderboards for each game and for full-trilogy runs. The first standard goals quickly stabilized around per-game Any% and completion-style categories, with the trilogy board also supporting full-compilation categories like Any% and 315%, plus platform distinctions that mattered for routing and consistency (for example, PC vs console, and even split definitions on PC where applicable).
Timing norms also hardened early, largely because load times can overwhelm “who played better” if they are left untreated. The trilogy’s leaderboard culture centers on recording both a real-time-with-loads value and a loadless value, with guidance for either subtracting standardized load totals or using a dedicated load-removal setup. That approach pushed runners toward common tooling, especially LiveSplit paired with a community load remover, so submissions could be compared on an even playing field.
As the game matured, post-launch changes created natural “eras” that the community preserved without turning the whole leaderboard into a patch-by-patch argument. Official updates added bonus levels like Stormy Ascent and Future Tense, which affected what “full completion” meant for some categories and helped justify keeping older runs in legacy lanes rather than rewriting history. The presence of legacy PS4 category groupings on the trilogy page reflects that philosophy: early standards were recorded, later standards were clarified, and both remained visible for historical continuity.
For N. Sane Trilogy, Speedrun.com functions as the public record and the front door. It is where categories live, where rules are displayed, and where the scene’s history stays browseable through leaderboards, guides, resources, forums, and streams, with a built-in “submit run” flow that keeps everything standardized.
Conversation and coaching tend to happen faster in Discord than in forum threads, and the community explicitly directs new runners there for day-to-day help, routing questions, and troubleshooting. The trilogy also benefits from being part of a wider Crash ecosystem, so you see both game-specific channels and broader series spaces used side by side depending on whether a question is “NST-only” or “Crash in general.”
Verification culture is shaped by two practical needs: proof and preservation. Runs are submitted with clear attribution and then marked as verified by the leaderboard staff once they meet the category’s rules and evidence requirements. Standards adapt when platforms change, too. For example, the trilogy community updated its submission guidance to avoid relying on Twitch highlights as durable proof, reflecting a broader emphasis on keeping record runs accessible long-term.
Knowledge is preserved in a very “speedrun-native” way: pinned tools, shared references, and reusable setups. The resources section collects community utilities, including timing and documentation aids, while technical threads and tool writeups help normalize best practices like load removal for consistent comparisons.
On Speedrun.com, Crash Bandicoot: N. Sane Trilogy is treated as a hub with full-game boards for each of the three games, plus “miscellaneous” boards that cover multi-game formats like Full Trilogy and Cross-Game. The main full-game categories reflect both fast finishes and full-completion goals. For Crash 1, the core set includes Any%, All Gems, 105%, and Secret Ending, with an additional rules split for “Major Glitches” versus “No Major Glitches.” For Crash 2, the headline categories are Any%, 100%, and 102%. For Crash 3, the core categories include Any%, a separate Any% No OoB option, and 108%.
A few policy decisions give the trilogy board its identity. Platform separation is built into the filters, typically distinguishing PC and Console for each game. Full Trilogy runs go further by separating PC runs based on whether the “File Fusion” approach is in play, alongside a Console option, and the community explicitly framed this as a leaderboard split rather than a blanket ban or an untracked technique. If runners want to use modifications that change the feel of the game, the community routes those to a separate Category Extensions board rather than the main leaderboards.
Timing rules are designed to reduce hardware and platform advantage from load screens. The leaderboards display both a loadless time and a time with loads, and community guidance treats loadless timing as the standard for comparison. The usual expectation is that runners use a LiveSplit load remover, with settings that also remove the black-screen transitions between loads. For submissions, proof requirements also matter, including a policy change that stops accepting Twitch highlights as evidence because they can disappear.
For the completion-style categories, the guiding principle is “reach the game’s maximum percentage,” even when that means the run is not defined by collecting every possible item. That philosophy is stated directly in rule discussions around 105% and 108%, where the community aligned requirements with what the game actually counts toward the final percentage rather than adding extra, arbitrary collectibles.
Because the trilogy remakes three different games under one package, strategy evolved along two tracks at once. Runners refined per game movement and level flow, while the broader community worked to standardize how the trilogy is practiced and timed. Early discussion often centered on how the remake’s feel differed from the original releases, and community threads comparing movement options helped establish what “fast” looked like in the remaster rather than assuming older techniques carried over unchanged.
As routing matured, the scene built a clear library of signature techniques for each title. In Crash Bandicoot 2: Cortex Strikes Back and Crash Bandicoot 3: Warped, “hobsliding” became one of the foundational movement skills runners are expected to learn, with dedicated tutorials that explain how to trigger it, maintain it, and apply it across stages. In Crash 3 specifically, route speedups also leaned into boundary and clip knowledge, including “gate clip” variants that were documented and organized into focused tutorial resources. In Crash Bandicoot, tech development is easy to see in the time trial side of the scene, where guides document setups like early finish methods and supporting movement requirements such as hyperjump execution.
Tooling did as much to improve consistency as any single skip. The community’s adoption of load removal was a turning point, especially across mixed hardware where loading behavior can distort comparisons. The trilogy scene maintains and shares load removal tooling, including a dedicated component that freezes timing through load screens. On PC, load remover and autosplitter functionality has been integrated into LiveSplit through the community setup notes, which lowered the barrier for new runners and made submissions more standardized. A parallel line of improvement came from memory based automation, with projects that auto start, auto split, and optionally surface in run metrics inside the timer itself.
The biggest strategy “paradigm shifts” tended to become rules conversations, not just route notes. The clearest example is File Fusion, a discovery that reshaped full trilogy routing enough that the scene formalized new category splits and clarified cross game boundaries, keeping both the classic format and the new approach legible on the leaderboard. Alongside that, practice quality improved through shared utilities, including community tools and selectors listed in the resources hub, which helped runners isolate techniques, rehearse difficult segments, and keep knowledge easy to find.
One of the first “scene-defining” milestones for Crash Bandicoot: N. Sane Trilogy was the moment the community’s record-keeping and category identity stabilized in one public home. With Speedrun.com serving as the central archive, the trilogy’s structure became legible to outsiders: distinct boards for each game, clear category families, and a shared expectation that runs are comparable because they are judged against the same published rules. That consolidation is what turns a set of fast playthroughs into a sport with standards.
A second major milestone type was timing normalization, because loading behavior can otherwise swamp performance. The adoption of loadless timing as a shared comparison standard, plus community tooling to support it, changed what “good” meant day to day: instead of arguing about hardware advantages, runners could focus on execution, route choice, and consistency. The release and continued refinement of a dedicated LiveSplit load remover, along with rules clarifying what counts as a load or transition, are the kind of infrastructure milestones that quietly reshape an entire leaderboard culture.
Another defining milestone type is the “route era split,” when a new technique or interpretation is treated as important enough to reshape the board rather than staying an informal footnote. In the trilogy, the formal split of Full Trilogy runs around “File Fusion” versus “No File Fusion,” alongside the introduction of cross-game formats, is a good example of a community choosing governance over chaos. Instead of endless debate about legitimacy, the leaderboard design itself preserved both styles and made the distinction readable.
Marathon showcases became their own watershed exposure moments, because they turned a niche leaderboard race into a performance narrative with commentary, explanations, and an audience that does not already know Crash tech. Runs appearing on Games Done Quick schedules and VODs helped codify the “watchable” form of the trilogy, especially for categories like Crash 3 Any% and full-trilogy showcases.
Finally, a modern milestone type is archival rigor: the scene adapting its evidence standards to keep history from disappearing. When Twitch introduced storage limits that threatened long-term availability of highlight archives, the trilogy community responded by updating submission proof expectations so record runs would remain accessible. That kind of policy shift does not change routes, but it changes what survives.
WingKn0x | Crash 1 (NST) Any% (MG) | 39:24 LRT (43:27 RTA) | Date: 1 year ago | PC | A clean benchmark example of the Major Glitches route, and a useful reference point when explaining why the scene tracks loadless time.
WingKn0x | Crash 1 (NST) Any% (NMG) | 41:56 LRT (46:07 RTA) | Date: 17 days ago | PC | A top-end “rules-tight” run that works well as a baseline for NMG expectations and consistency.
BustaCarl | Crash 1 (NST) Any% (NMG) | 44:44 LRT (49:14 RTA) | Date: 4 years ago | PC | An older-era NMG performance; the page is flagged “Video at risk,” making it a good candidate to archive off-site for historical recordkeeping.
Krisuchu | Crash 2 (NST) Any% | 39:15 LRT (43:16 RTA) | Date: 8 months ago | PC | A modern benchmark run that’s handy for describing Warp Room flow and cumulative micro-saves across transitions.
CasuallyClutching | Crash 3 (NST) Any% | 37:36 LRT (41:23 RTA) | Date: 1 month ago | PC | A strong reference for the “open-tech” Any% ruleset, especially when you later contrast it with No OoB.
Alaapo | Crash 3 (NST) Any% (No OoB) | 41:41 LRT (46:34 RTA) | Date: 2 years ago | PC | A representative No OoB benchmark that illustrates how the route stays grounded while still emphasizing execution and sequencing.
CasuallyClutching | Full Trilogy Any% (No File Fusion) | 2:03:56 LRT (2:16:45 RTA) | Date: 10 months ago | PC (No File Fusion) | A “mainline” trilogy Any% reference that keeps the run honest by excluding File Fusion warps, while still rewarding long-form pacing and consistency.
CasuallyClutching | Full Trilogy Any% (File Fusion) | 9:29 LRT (9:58 RTA) | Date: 11 months ago | PC (File Fusion) | A clear demonstration of why File Fusion is treated as its own category: it compresses the trilogy into a radically different, warp-heavy execution test.
McCrodi | Full Trilogy Any% (File Fusion) | 10:30 LRT (11:01 RTA) | Date: 4 years ago | PC (File Fusion) | A historically useful older File Fusion reference point, also flagged “Video at risk” for preservation awareness.
CookieNaval | Full Trilogy 315% | 5:57:52 LRT (6:52:44 RTA) | Date: 2 years ago | PC (No File Fusion) | The current top completionist benchmark for the trilogy’s full objective set, and a great pacing/stamina reference for marathon-length runs.
Dan Tanguay. “Previously Unreleased Stormy Ascent Level Out Today for Crash Bandicoot N. Sane Trilogy.” PlayStation.Blog, July 20, 2017. https://blog.playstation.com/2017/07/20/previously-unreleased-stormy-ascent-level-out-today-for-crash-bandicoot-n-sane-trilogy/
Kyle Martin. “Crash Bandicoot N. Sane Trilogy Gets New Level, HDR, Faster Loading.” PlayStation.Blog, June 29, 2018. https://blog.playstation.com/2018/06/29/crash-bandicoot-n-sane-trilogy-gets-new-level-hdr-faster-loading/
Activision Games Blog. “The Crash Bandicoot N. Sane Trilogy is Arriving Early.” May 10, 2018. https://blog.activision.com/crash-bandicoot/2018-05/The-Crash-Bandicoot-N-Sane-Trilogy-is-Arriving-Early
Activision Games Blog. “Wumpa Fruit for Everyone! Crash Bandicoot N. Sane Trilogy Makes Its Way to Nintendo Switch, Xbox One, and Steam for the First Time.” June 29, 2018. https://blog.activision.com/crash-bandicoot/2018-06/Wumpa-Fruit-for-Everyone-Crash-Bandicoot-N-Sane-Trilogy-Makes-Its-Way-to-Nintendo-Switch-Xbox-One-and-Steam-for-the-First-Time
Steam Store. “Crash Bandicoot N. Sane Trilogy.” Accessed February 11, 2026. https://store.steampowered.com/app/731490/Crash_Bandicoot_N_Sane_Trilogy/
Speedrun.com. “Crash Bandicoot: N. Sane Trilogy.” Accessed February 11, 2026. https://www.speedrun.com/crashnst
Speedrun.com. “Full Trilogy, Any%, PC (No File Fusion) Leaderboard (rules and submissions).” Accessed February 11, 2026. https://www.speedrun.com/crashnst?h=Full_Trilogy-any-pc-no-file-fusion&x=wkpql9vd-j84mmj89.81p4yo81-5lypq92l.z1979r4l
Speedrun.com Forums. “Loading Times, Info for Submissions.” Accessed February 11, 2026. https://www.speedrun.com/crashnst/forums/117vl
Speedrun.com Forums. “Load Remover Tool for LiveSplit.” Accessed February 11, 2026. https://www.speedrun.com/crashnst/forums/qqhrs
Speedrun.com Forums. “Big Discovery: File Fusion, and Cross-Game Runs.” Accessed February 11, 2026. https://www.speedrun.com/crashnst/forums/0kde5
Speedrun.com. “Crash Bandicoot: N. Sane Trilogy.” Accessed February 11, 2026. https://www.speedrun.com/crashnst
Jess Weatherbed. “Twitch Is Limiting Streamers to 100 Hours of Highlights and Uploads.” The Verge, February 20, 2025. https://www.theverge.com/news/616296/twitch-storage-cap-highlights-uploads-speedrunning
Speedrun.com Guides. “All Games, Installing the Load Remover for PC.” Accessed February 11, 2026. https://www.speedrun.com/crashnst/guides/6co5d
Speedrun.com Guides. “Crash 3, Any% Tutorial.” Accessed February 11, 2026. https://www.speedrun.com/crashnst/guides/svtav
Speedrun.com Guides. “All Games, Installing the Load Remover for Console.” Accessed February 11, 2026. https://www.speedrun.com/crashnst/guides/6lita
Thomas Neff. “LiveSplit.CrashNSTLoadRemoval.” GitHub repository. Accessed February 11, 2026. https://github.com/thomasneff/LiveSplit.CrashNSTLoadRemoval
Thomas Neff. “Releases: LiveSplit.CrashNSTLoadRemoval.” GitHub. Accessed February 11, 2026. https://github.com/thomasneff/LiveSplit.CrashNSTLoadRemoval/releases
Grimelios. “LiveSplit.Crash (Memory-Based Autosplitter).” GitHub repository. Accessed February 11, 2026. https://github.com/Grimelios/LiveSplit.Crash
Games Done Quick. “Schedule: Summer Games Done Quick 2023.” Accessed February 11, 2026. https://gamesdonequick.com/schedule/43
Games Done Quick. “Crash Bandicoot: N. Sane Trilogy by Murcaz in 2:26:50 (Summer Games Done Quick 2023).” YouTube video. Accessed February 11, 2026. https://www.youtube.com/watch?v=2EYUC5R7m7E
GDQ VODs. “Crash Bandicoot: N. Sane Trilogy (PC).” Accessed February 11, 2026. https://gdqvods.com/game/crash-bandicoot-n-sane-trilogy-pc
Joshua Wolens. “Crash Bandicoot’s Original Lead Programmer Reckons the Overhauled N. Sane Trilogy Got Almost Everything Right Except for One Tiny Detail: ‘They Completely Botched How Jumping Works’.” PC Gamer. Accessed February 11, 2026. https://www.pcgamer.com/games/crash-bandicoots-original-lead-programmer-reckons-the-overhauled-n-sane-trilogy-got-almost-everything-right-except-for-one-tiny-detail-they-completely-botched-how-jumping-works/