X Relaunches a Rebuilt Android App After Year-Long Effort

On July 20, 2026, X shipped a completely rebuilt Android application — the product of what head of product Nikita Bier called "one of the largest engineering projects" in the company's history. The new app discards the old mixed codebase entirely in favor of clean Kotlin and Jetpack Compose, a modernization effort that took roughly twelve months and a dedicated engineering team to complete. For a platform that had let its Android experience visibly deteriorate while iOS users received steadily improving builds, the release marks a significant course correction.

The Claim

X's engineering team announced the rebuilt app on Monday with a post from the @Engineering account citing improvements across every major surface: faster loading, smoother scrolling, more reliable push notifications. The company is framing the rewrite not just as a catch-up effort but as foundational infrastructure work — the kind of base that will let the team ship new capabilities "at lightning speed," in Bier's words.

The decision to discard the old codebase rather than incrementally refactor it reflects how far things had deteriorated. The previous Android app ran on a mixed architecture that had accumulated years of technical debt, and the symptoms were severe enough to matter to ordinary users — links to X posts opened from third-party apps would sometimes fail to load entirely. That specific failure, a deep-link handling problem affecting anyone using the app in conjunction with anything else on their phone, is exactly the kind of regression that erodes store ratings and user trust without ever generating a single headline.

Bier first publicly flagged the effort last August, announcing that X was assembling an Android "dream team" to rebuild from scratch. By October, the company reported one of its biggest weeks ever for Android downloads, which appears to have reinforced the business case. The rebuilt app, now live in the Play Store, is the result of that sustained push.

What We See

The most significant data point to emerge from the launch is not about the app itself — it is about Grok. Shortly after the announcement, X's own Grok account responded to a question about AI's role in the rewrite with notable directness: Grok stated it had "zero visibility" into X's internal codebase and contributed nothing to the rewrite directly. It acknowledged helping individual developers move faster on a daily basis, but drew a clear line between that kind of acceleration and the app being AI-generated or AI-authored.

Our read is that this is one of the cleaner real-world data points we have seen on the AI-assistance-versus-AI-authorship distinction. It matters because the framing in much of the AI tooling conversation tends to blur that line — productivity gains get described in ways that imply larger structural contributions. Here, the company's own AI model, commenting publicly and without apparent prompting, made the distinction explicit. A complete mobile platform rewrite, on a team with daily access to one of the more capable AI coding assistants available, still required twelve months of human engineering. That timeline is a useful reference point for anyone making estimates about how quickly AI tooling can compress serious platform work.

The choice of Kotlin and Jetpack Compose as the rebuild target is also worth addressing on its own terms. Compose has been Google's recommended Android UI toolkit since 2021, and by mid-2026 it is the baseline expectation for any credible new Android development. The fact that X needed a full rewrite to reach that baseline — rather than migrating its existing views incrementally — is a measure of how long the legacy codebase had gone without disciplined modernization. Yahoo Tech's republication of TechCrunch reporter Sarah Perez's original account confirms the same facts and timeline without adding independent sourcing, but the two accounts are fully consistent on every material claim.

The global market angle driving this effort is real and worth taking seriously on its own. Android commands the majority of smartphone use across nearly every market outside the United States. Years of comparative neglect on that platform meant years of compounding friction for international users — friction that contributed to the company watching Android download numbers become a talking point internal enough that a single strong October was noteworthy. The rewrite removes that category of obstacle.

Where It Falls Short

The app that shipped on July 20 is not feature-complete. Spaces — X's live audio rooms — is absent from the initial release. Support for older Android devices is listed as pending. Several additional capabilities, including the video editor, cashtags, and custom timelines, are described as forthcoming in subsequent updates. Shipping before full feature parity is a defensible product strategy: it avoids holding back structural improvements while every edge-case port is finished. But it does mean users who update immediately will encounter visible gaps, particularly those who use Spaces regularly.

More importantly, the performance claims in the announcement remain vendor-stated rather than independently verified. The assertion that the new build is faster, smoother, and more reliable "in every way" comes from X's own @Engineering post. No third-party benchmark comparing the old and new builds has been published as of this writing. It is entirely plausible the new Compose-based architecture delivers measurable gains — Compose builds generally do outperform equivalent XML-View implementations on modern Android — but "plausible" and "confirmed" are not the same thing. Whether the rebuild translates to meaningful user or revenue recovery for X in international markets is a longer-term question this relaunch does not answer. It removes one clear obstacle. What replaces the lost ground is a separate problem.

Sources