Claude by Anthropic: Apps, Plans and Models Explained (2026)

Claude by Anthropic is an AI assistant made by the AI company Anthropic, available free on the web at claude.ai, as desktop apps for Mac, Windows and Linux (beta), and as mobile apps. “Claude by Anthropic” is also the exact name of the official iPhone app in the App Store, which helps explain why so many people are searching for it right now.

Below we cover what Claude is, how to install it safely, what each plan costs, and what changed with the new Claude 5.5 models released over the past few weeks. Everything here reflects Anthropic’s own pages as of October 13, 2026; plans and features change often, so check the linked pages before you pay for anything.

What Claude actually is

Claude is both a family of AI models and the chat app built on them. In the app you can talk to Claude, upload files, search the web, and have it create documents and run code. Developers can also use the same models through Anthropic’s API and cloud partners such as Amazon Web Services, Google Cloud and Microsoft Azure.

According to Anthropic’s pricing page, even the Free plan includes:

• Chat on web, desktop and mobile
• Web search
• File creation and code execution
• Memory
• Artifacts (interactive outputs such as small apps and pages)
• Claude Design, Slides and Docs

Paid plans add things like Claude Code (Anthropic’s coding agent), Projects, Research, Claude in Chrome and Microsoft 365 integration, and task scheduling.

How to install the official Claude app

Only download Claude from Anthropic’s own download page or the official app stores. Search results for “Claude download” include low-quality copycat pages; ignore them.

iPhone and iPad

1. Open the App Store.
2. Search for “Claude by Anthropic”.
3. Check that the seller is Anthropic, then tap Get.
4. Open the app and sign in or create an account.

Anthropic’s Help Center says the app requires iOS 18.0 or iPadOS 18.0 or later. There is also a separate “Claude for Intune” app meant only for Enterprise customers whose IT team manages devices with Microsoft Intune. If you’re not in that situation, install the standard app.

If the app doesn’t appear, Anthropic lists three likely reasons: an unsupported region, an unsupported device, or an older iOS version.

Android

Anthropic’s Help Center has a dedicated Claude for Android section, and the Android app is distributed through Google Play. Note that the current download page highlights the iOS app and points other phone users to claude.ai/mobile, so the easiest route on Android is to start from that link and follow the official Play Store listing it leads to.

Desktop

The download page offers:

The desktop app adds a quick-entry hotkey, desktop extensions that connect Claude to local files and apps, Claude Code, and Cowork, a mode for working on tasks with your local files. On Linux, Claude Code and Cowork need a paid plan, and Computer Use and Dictation are not yet available.

Claude plans and prices (October 2026)

Here is what Anthropic’s pricing page lists today. Prices are in US dollars.

A few details worth knowing:

• Max includes higher output limits, priority access at busy times, early access to new features, and $100 to $200 a month in API credits depending on the tier.
• Team includes central billing, SSO, admin controls, and Anthropic states it does not train models on your content by default. It also includes up to $500 a month in pooled API credits.
• Education plans exist, but the pricing page doesn’t list a price.

When Opus 5.5 launched, Anthropic also said it was raising five-hour usage limits on Pro, Max, Team and seat-based Enterprise plans. Exact usage limits aren’t published as fixed numbers, so expect them to vary.

The new Claude 5.5 models

Anthropic released a new generation of models in quick succession:

The pricing page lists Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5 as the current models, with older ones such as Opus 5 and Sonnet 5 marked as legacy. A “Mythos” entry also appears in the site’s model menu, but we couldn’t find official details on it.

Opus 5.5

Anthropic’s announcement says Opus 5.5 produces output more than 30% faster than Opus 5 and costs about 40% less to run on typical workloads at default settings. Anthropic also says it is more resistant to prompt injection than Opus 5 and writes more clearly, putting key information first. Its thinking mode can’t be switched off.

Sonnet 5.5

According to the Sonnet 5.5 page, it generates output 30%+ faster than Sonnet 5. In the Claude apps and Claude Code, its default effort level is set to Medium, and you can adjust effort to trade speed and cost against quality.

Haiku 5.5

The Haiku 5.5 page describes it as built for high-volume tasks such as summaries, classification and database queries, and fast enough for live customer support. It is the first Haiku model with an adjustable effort setting.

One thing Anthropic’s announcements don’t spell out is exactly which models each app plan can pick from. Check the model selector inside the app after you sign in.

What it costs developers

If you use Claude through the API rather than the app, Anthropic lists these prices per million tokens:

Haiku 5.5 is priced higher for prompts over 100k tokens ($0.50 input, $2.50 output). App subscribers don’t pay these rates directly.

Is Claude worth trying?

If you want a general AI assistant, the Free plan is enough to judge whether Claude suits your writing, research or coding style, and it already includes memory, web search and file creation. Upgrade to Pro if you hit usage limits regularly or want Claude Code and Research. Max mainly makes sense if you use Claude for hours a day, especially for coding.

Two practical tips:

1. Install only from the App Store, Google Play, the Microsoft Store or claude.com, and never enter your login on a third-party “Claude download” site.
2. If you subscribe through the app and later delete it, remember that removing the app does not cancel your subscription.

All figures above are as of October 2026 and come from Anthropic’s own pages. Anthropic updates models and plans frequently, so treat prices and limits as a snapshot rather than a guarantee.