Nitter Is Dead (Again): Inside X's Crackdown

TL;DR

• X Corp shut down Nitter and XCancel a second time on September 14-15, 2026, after both had returned online on September 7 following legal advice.
• The original cease-and-desist letters (August 24, 2026) accused Nitter of unlawfully circumventing X's API and accessing account and session data, citing the Texas Harmful Access by Computer Act and the Lanham Act.
• There's no free, login-free replacement: the route X itself permits is its paid API, which now has no free tier and jumps to $5,000+/month for meaningful volume.
• The shutdown lands weeks after X quietly closed its own API's flat-rate self-serve tiers too, a second, unrelated tightening of the same door.

What happened to Nitter and XCancel

On August 24, 2026, X Corp sent cease-and-desist letters to Nitter's maintainer and to XCancel, demanding a permanent takedown by 5 pm EST the next day. Both complied, and nearly every popular Nitter instance went dark within hours. Per TechCrunch's reporting on the letter's contents, X accused Nitter of "unlawful use and circumvention" of its API, alleging the project accessed X accounts and session tokens in violation of X's rules. The letters cited, among others, two specific statutes: the Texas Harmful Access by Computer Act, a civil cause of action for knowing or intentional computer-access violations, and the Lanham Act, the federal trademark statute.

On September 7, Nitter announced it was continuing, "following legal advice," and came back online alongside XCancel. That held for four days. Nitter's GitHub repository was archived to read-only again on September 11, and XCancel suspended service a second time on September 14, citing "a new development in the ongoing legal proceedings."

This isn't Nitter's first brush with X's defenses. A 2024 lockdown took the flagship instance offline temporarily through purely technical means, API restrictions rather than legal pressure, and development resumed once that pressure eased. This time is different on both counts: the approach is legal, not technical, and after a second shutdown following a second return, it looks considerably more final, one piece of a broader pattern of platforms locking down public data access. The story became one of the most-discussed threads in recent memory, a sign of how many people had quietly built the tool into their daily reading.

Why Nitter mattered, and why there's no direct replacement

Nitter's whole pitch was subtraction. It let anyone read public X posts, threads, and profiles without an account, without the app, without ads, without tracking, and without JavaScript. No home feed, no notifications, no posting, nothing beyond a clean, read-only mirror of whatever was already public.

It ended up embedded far beyond a browser tab. Wikipedia citations linked to Nitter mirrors instead of X directly. RSS feeds ran on it to turn accounts into subscribable updates. Entire online communities adopted a rule of posting only Nitter links, never direct X links, treating the swap itself as a small, deliberate protest against Elon Musk's ownership of the platform.

Is there a Nitter alternative?

No, not one that matches what Nitter actually offered. Its method, standing between a reader and X's public pages without either party needing an account, is exactly what X's legal action targeted, so anything built the same way inherits the same risk. That leaves three realistic paths, not one.

The first is X's own API, now fully paid with no free tier, which the next section covers in detail. The second is simply using X.com itself, account, ads, tracking, and all, the tradeoff Nitter existed specifically to avoid. There's also the option to buy the Premium subscription to avoid the ads and get a better experience. The third applies more broadly than just X: anyone whose research, product, or monitoring workflow leaned on this kind of open, no-account access isn't only facing a Nitter-shaped hole. That's a dependency worth rethinking generally, not patching around this one loss, and it means moving off ad hoc workarounds onto infrastructure actually built for compliant, ongoing collection across many sources, not just the one that just disappeared. More on what that looks like, concretely, later in this piece.

The official X API: What it costs now

X's own API is the path X itself permits: no scraping, no session tokens, no gray area, just a metered, paid interface X controls directly. As of 2026, that compliance comes at a real cost.

The standalone free tier is gone. New developers default to pay-per-use pricing, somewhere between $0.001 and $0.20 per read or post action depending on the type, capped at roughly 2 million reads a month before Enterprise pricing takes over. The older flat-rate tiers, Basic at $200 a month and Pro at $5,000 a month, are closed to new signups entirely, and existing Pro subscribers are reportedly being migrated off the flat rate altogether. Enterprise access is a custom deal, reportedly starting around $42,000 a month.

For compliance-sensitive or genuinely low-volume work, this is the low-risk call, and the only one that doesn't carry legal risk. For anyone hoping to replicate what Nitter did at real scale, or for free, it's a non-starter. Choosing between an API and other data-collection approaches is its own decision with its own tradeoffs.

What locked-down platforms mean for your data strategy

Two things happened within weeks of each other in 2026, and neither is really about Nitter specifically. Nitter and XCancel got shut down through legal action. Around the same time, X finished retiring its own API's free and flat-rate self-serve tiers, a separate, unrelated tightening of the same door. Neither is an isolated event either. Reddit's 2023 API pricing change ended most third-party apps overnight in exactly the same way, with exactly the same lack of warning. Platforms generally have gotten more aggressive about controlling who reads their public data and how, whatever the platform happens to be.

That's the actual story here, bigger than any one shutdown. Open access to platform data is narrowing across the board. Any workflow, product, or research pipeline quietly built on one platform's willingness to look the other way is exposed to exactly this kind of overnight change, with no notice and no appeal.

Decodo's Web Scraping API and residential proxies exist for the sites that remain open to compliant, public-web collection, infrastructure built for that job across many sources, not a workaround for the platforms actively fighting to close it. Decodo also co-founded the Ethical Web Data Collection Initiative, an industry effort specifically aimed at keeping that kind of collection trustworthy and responsible rather than adversarial.

Is it legal to scrape public X data?

U.S. law generally treats publicly available web data differently from data sitting behind a login. The standard reference point is hiQ Labs v. LinkedIn, where accessing data anyone could view without logging in was found not to violate the Computer Fraud and Abuse Act, since nothing was actually being "hacked" to get to it.

That's not the whole picture, though. A platform's Terms of Service is a separate, contractual matter, independent of criminal law entirely. Violating it can still mean an account ban, an IP ban, or civil exposure, even in a case where no criminal statute is implicated at all. Being on the right side of the CFAA doesn't mean a platform has no recourse.

The practical guidance: stay scoped to data that's genuinely public, nothing behind a login. Avoid anything resembling account or session-token access. Respect published rate limits rather than treating them as a suggestion. And if you're storing EU users' public post data, GDPR still applies to how you handle and retain it. Publicly posted isn't the same as unregulated.

Worth being precise about where X's actual allegations against Nitter sit in this picture, since it matters for weighing your own risk: X accused Nitter of API circumvention and session-token access, conduct well beyond simple public-page scraping. That's a meaningfully different, more aggressive category of behavior than reading pages anyone could already see in a browser, and conflating the two would be a mistake in either direction.

None of this is legal advice, and the specifics vary by situation. Our guide to web scraping legality goes deeper on the cases, the statutes, and where the real lines tend to fall.

Final thoughts

For now, Nitter and XCancel's second shutdown closes the last free, no-account way to read public X data, though the brief return in between is reason enough not to call this permanent with total confidence. X tightened its own API pricing on almost the same timeline, a separate move that points to the same underlying shift. Whatever your data strategy actually depends on, it needs a real answer for a platform changing the rules overnight with no warning. For teams building that answer now, Decodo's Web Scraping API and residential proxies are built for exactly that kind of resilience.