How to switch between multiple Claude Code accounts in 2026: 6 ways compared

Disclosure: Enjoy is our product, and it keeps several Claude Code accounts signed in. We checked every way below against Anthropic’s documentation and each maker’s own pages on October 31, 2026, and we say where another way is the better pick, including the free ones that need nothing but Claude Code.

The ways to use more than one Claude Code account, compared

People keep a second Claude account for one of two reasons: a work plan beside a personal one, or a spare for when the first runs out. Claude Code keeps one login per configuration directory, so the questions are how you switch, whether both accounts stay signed in at once, and whether anything switches for you. Anthropic’s authentication guide gives its own answer: “To stay signed in to multiple accounts at once, such as work and personal accounts, give each account its own configuration directory.”

Way	Best for	Price, checked October 31, 2026	Signed in at once	Switches for you	Runs on
/login	Changing accounts now and then	Free with Claude Code	One	No	Mac, Windows, Linux
A configuration directory per account	A work and a personal account, side by side	Free with Claude Code	One per directory	No	Mac, Windows, Linux
A token per account	Scripts, CI and scheduled jobs on another account	Free with Claude Code	One per job	No	Wherever Claude Code runs
claude-swap	Switching from a terminal, and rotating before a limit	Free, open source	Several, with cswap run	Yes, at 90% of a limit by default	Mac, Windows, Linux, WSL
clauth	A terminal dashboard with a fallback chain	Free, open source	Several, with clauth start	Yes, down a chain you set	Linux, Mac; Windows through Git Bash or MSYS2
Enjoy	Every account signed in, one per project, without a terminal	A monthly subscription, free for 7 days, with your phone and other computers (pricing)	Every account you add	Yes, when one runs out	Mac, Windows, Linux

Every account is a Claude plan of its own. Claude’s pricing has Pro at $20 a month, or $17 a month billed yearly, Max at $100 and Ultra at $200, each with “Claude Code included,” and the Free plan leaves Claude Code out.

Before you start: what’s allowed, and which account is in use

• More than one account is allowed; sharing one isn’t. In February 2026, Thariq, of Anthropic’s Claude Code team, wrote that “it’s not against terms of service to have multiple MAX accounts,” and that using them to resell tokens is. Anthropic’s Consumer Terms say “You may not share your Account login information, Anthropic API key, or Account credentials with anyone else,” and Claude Code’s legal page says “Advertised usage limits for Pro, Max, and Ultra plans assume ordinary, individual usage of Claude Code and the Agent SDK.” Our help guide covers what does get accounts suspended.
• Check which account is in use. /status shows “version, model, account, and connectivity.” An API key outranks every login: “If you have an active Claude subscription but also have ANTHROPIC_API_KEY set in your environment, Claude Code uses the API key once you approve it.”
• A work computer may decide for you. An employer can set forceLoginOrgUUID in managed settings, and then “Claude Code reports an error for a claude.ai login to any other organization.” Whether a personal account belongs on a work machine at all is a question for your employer; our memo on the Accenture report shows why it matters.
• Some tools keep copies of your login. Claude Code’s legal page says developers “may not collect, store, or intermediate Claude.ai credentials or session tokens — sign-in to a Claude account must complete through Anthropic’s own flow.” claude-swap and clauth keep copies of each account’s tokens on your computer, and Anthropic hasn’t said whether that line covers a tool you run on your own accounts.

/login, one account at a time

Best for: changing accounts now and then, such as moving to your work plan for the afternoon.

Claude Code’s commands list /login as “Sign in to your Anthropic account.” Run it in a session, sign in with the other account in your browser, and Claude Code switches to it; its docs describe this as when “you switch accounts with /login.” Your settings and session history stay as they were, because they live in the configuration directory, not with the login.

Price: free with Claude Code.

Watch for: there’s one login at a time, so each switch means signing in again. Switching ends Claude Code’s wait for a limit to reset, so send a message to carry on. And /logout “also resets your first-launch setup state,” so you go through setup again. On claude.ai itself, Anthropic’s help center switches between an individual account and a Team or Enterprise account “tied to the same email address” from your initials in the corner, but it doesn’t cover Claude Code.

A configuration directory per account

Best for: a work and a personal account, both signed in, with nothing to install.

This is Anthropic’s own way. Each directory “has its own settings, session history, and claude.ai login or API key,” and you pick one with the CLAUDE_CONFIG_DIR environment variable when you start Claude Code. The docs’ example, for Bash or Zsh, goes in ~/.bashrc or ~/.zshrc:

alias claude-work='CLAUDE_CONFIG_DIR=~/.claude-work claude'

Then claude-work uses your work account while claude keeps your personal one. “After you open a new terminal and run claude-work for the first time, Claude Code walks you through login and setup for the new directory.” On a Mac, Claude Code “keys the macOS Keychain entry to that directory too, so a session with a different CLAUDE_CONFIG_DIR reads a different entry.” Two terminals can run two accounts at the same time.

Price: free with Claude Code.

Watch for: everything in the directory is separate. “All settings, session history, and plugins are stored under this path,” per the environment variables reference, so your user settings, plugins and conversations don’t carry from one account to the other, and you set up each directory on its own. The variable is “Ignored in project and local settings,” so a repository can’t choose the account for you. And “Separate directories don’t keep two Claude Console sign-ins without an API key apart, because Claude Code stores that kind of sign-in outside the configuration directory.”

A token per account, for scripts

Best for: scripts, CI and scheduled jobs that should run on a different account from the one you type in.

claude setup-token “opens the same browser authorization flow as /login,” then prints a one-year token instead of saving it. Sign in with the account the job should use, and set the token as CLAUDE_CODE_OAUTH_TOKEN in that job’s environment. It “Takes precedence over keychain-stored credentials,” so the job runs on that account while your own terminal stays on yours.

Price: free with Claude Code. The token “authenticates with your Claude subscription and requires a Pro, Max, Ultra, Team, or Enterprise plan.”

Watch for: “It can only make model requests, so it can’t establish Remote Control sessions or fetch claude.ai connectors,” and bare mode “does not read CLAUDE_CODE_OAUTH_TOKEN.” The variable sticks: if you run /login while it’s set, Claude Code switches that session but “reads the variable again in every new session until you remove it from your shell profile.” Treat the token as a password, since the Consumer Terms forbid sharing your account credentials.

claude-swap

Best for: switching accounts from a terminal without signing out, and letting a tool rotate them before you hit a limit, free.

claude-swap is a “Multi-account switcher for Claude Code,” free under the MIT license, with more than 4,000 stars on GitHub. It installs with uv tool install claude-swap or pipx install claude-swap, needs Python 3.12 or later, and “Works with both the Claude Code CLI and the VS Code extension.” Sign in to each account in Claude Code and run cswap add, then cswap switch moves to the next account, or cswap switch 2 to a chosen one. cswap auto switches “When the active account’s 5-hour or 7-day window reaches the threshold (default 90%),” and cswap run 2 launches Claude Code “as a specific account in the current terminal only,” so two accounts can work at once. cswap map binds a directory to an account, such as your work repositories to your work plan.

Price: free.

Watch for: it works by keeping copies of each account’s login: it “Backs up OAuth tokens and config when you add an account,” in the macOS Keychain or in files on Windows and Linux. Its README warns, “Do not run /logout first: current Claude Code may revoke the refresh token stored for the account you are leaving.” On a Mac, a running session picks up a switch once Claude Code’s Keychain cache, about 30 seconds, expires. Accounts started with cswap run keep their own chat history unless you pass --share-history, and after a switch, “the first message on the new account may use extra usage as its conversation cache rebuilds.” Its menu bar app is for macOS only.

clauth

Best for: a terminal dashboard of every account’s usage, with a fallback chain that moves you on before a limit stops you.

clauth is a “Claude Code multi-account manager, usage monitor (CLI, TUI & cross-account delegation via MCP),” free under the MIT license on GitHub, and its latest version on docs.rs, 0.18.0, came out on October 25. It installs with cargo install clauth or a script, on “Linux, macOS, Windows (Git Bash / MSYS2).” clauth capture work saves the login you’re in as a profile, clauth work switches to it, and clauth start personal runs Claude Code under another account “in its own config dir,” so several run at once. It can “Auto-switch down a fallback chain the moment an account hits its limit,” and its plugin lets a live session “list, switch, or delegate a whole prompt (even headless) to another account.” It handles Codex logins too.

Price: free.

Watch for: clauth “handles live OAuth tokens and replaces its own binary over the network,” its README says, and keeps them “under ~/.clauth/, with 0600 permissions on Unix.” Binary installs “update themselves in the background,” checksum- and signature-verified, which CLAUTH_NO_UPDATE=1 turns off. It’s a smaller project than claude-swap, with fewer than 300 stars, and on Windows it needs Git Bash or MSYS2.

Enjoy

Best for: keeping every Claude Code account signed in, one per project, and moving a conversation to one with usage left, without a terminal.

Enjoy is a desktop app for Mac, Windows and Linux that runs Claude Code, Codex, Grok Build, Antigravity, Muse Code, and local models through OpenCode, on the plans you already have. Its saved accounts “stay signed in side by side, so you can switch without signing out,” and each one signs in “through the agent itself, on Anthropic’s or OpenAI’s own sign-in page, the same way it would in a terminal.” What it adds:

• An account per project. A work project stays on your work plan and a personal one on yours. One of the accounts is the sign-in Claude Code also uses in your terminal.
• Usage side by side. Each signed-in account shows its five-hour and weekly bars and when each resets, so you can compare before you choose.
• It switches when one runs out. Enjoy “switches this project to the first one that still has usage, and the conversation continues there on its own,” or, if you’d rather, waits for the reset and carries on.
• The conversation comes along. Claude Code accounts don’t share conversations, so when one continues on another account, “Enjoy first copies Claude’s history of it to that account, so Claude carries on where it left off.” Saved accounts work for Codex and Grok Build too.

Price: Enjoy is one monthly subscription, and the first seven days are free, with a card to start. It includes your phone and any browser and your other computers, and each teammate you invite needs a subscription of their own. See pricing.

Watch for: Enjoy doesn’t change Anthropic’s limits, and each account is a plan you pay for. On Windows, saved accounts need Developer Mode turned on, “except in projects inside WSL.” A key set up for the agent on your computer “overrides every account, so saved accounts can’t separate them.” Each Claude Code account “keeps its own notes about your projects and its own plans from Plan mode.” Waiting for a reset needs your computer on, with Enjoy open, when the usage resets. Enjoy isn’t open source, it’s a monthly subscription on top of your Claude plans once the free week ends, and scheduled workflows run only while your computer is awake with Enjoy open.

When one account is enough

If the second account is only for running out, one bigger plan may be simpler, and by Anthropic’s own multiples it costs about the same: two Pro plans are $40 a month for twice Pro’s usage, Max is $100 for “5x Pro usage,” and Ultra is $200 for “10x Pro usage every week,” with “No 5-hour limit” for verified accounts. One account means one set of settings, one history and nothing to switch. Our Claude Pro vs Max vs Ultra guide compares the plans, and our usage limit guide covers waiting, usage credits and the other ways through a limit.

Keep two when they really are two: an employer’s plan and your own. Claude Code’s legal page puts Team and Enterprise under Anthropic’s Commercial Terms, and Free, Pro, Max and Ultra under its Consumer Terms, so work you do for an employer belongs on the account it pays for. For that, a configuration directory per account is free and takes one line.

How we compared

We read Anthropic’s Claude Code docs on authentication, commands, environment variables and interactive mode, its legal page, Consumer Terms, help center and pricing, Thariq’s post, OpenAI’s help center articles on device limits and multiple accounts, and claude-swap’s and clauth’s READMEs and listings on October 31, 2026, and every one is linked below. Licenses and star counts come from GitHub that day. We quote each maker’s own words for its limits and leave out what it hasn’t documented, and we didn’t test the community tools ourselves. Claude Code’s sign-in changes often, so we recheck this guide against its sources at least once a month, and the date at the foot of the page says when we last did.