Do Claude chat and Claude Code share the same limits?

Yes. Chatting on claude.ai, using Claude Desktop or the mobile apps, and running Claude Code all count against the same plan limits. Anthropic's wording: "Note that your usage of all different Claude product surfaces (claude.ai, Claude Code, Claude Desktop) counts towards the same usage limit" (Help Center).

So a long chat leaves less for Claude Code in the same five-hour session and the same week, and the other way round.

What is shared

For Pro, Max and Ultra plans, Anthropic says the limits "are shared across Claude and Claude Code, meaning all activity in both tools counts against the same limits". That covers the five-hour session limit, the weekly limits and, where your plan has one, the separate weekly limit for Fable (Help Center, as of November 2026).

Claude Code inside editors is included too. Claude Code in VS Code, Cursor and other VS Code forks, and JetBrains IDEs is covered by your plan, and "IDE usage counts toward the same usage limits shared across Claude and Claude Code" (same article). If you want the editor angle in detail, see do Claude Code limits apply in VS Code and JetBrains.

Claude Desktop has a "Code" tab that includes Claude Code and uses the account signed in to the app (Help Center). Whatever you do there comes out of the same allowance.

Non-interactive use counts as well. When you are signed in with a Claude plan, claude -p and the Agent SDK "still draw from your plan's usage limits", and runs started by the Claude Code GitHub Action, an IDE extension or the Claude desktop app "count as Claude Code usage" (Help Center). More on that in does claude -p use your subscription limits.

One account, not one device

The limits belong to your account, not to a computer. Use Claude on your phone, in a browser and in a terminal, and all of it lands in the same place. That has one practical consequence: the Claude Code /usage breakdown is computed "from local session history on this machine, so usage from other devices or claude.ai is not included" (Claude Code docs).

The breakdown of what used your limit (skills, subagents, MCP servers and so on) only knows about this Mac. The plan limit bars are account-wide, though, and they include everything. So the bars can be high while the breakdown on this machine looks small, and if you also use Claude elsewhere, that gap is expected. How to check your Claude Code usage with /usage explains the screen.

The exception: Claude Code-only changes

Shared limits do not mean every change applies to every product. From May 13 through September 13, 2026, Anthropic raised the weekly limit in Claude Code by 50% as a promotion. It "applied to Claude Code only ... Usage limits for other Claude products, like Claude (web, desktop, and mobile) and Claude Cowork, were unchanged" (Help Center). Since September 14, 2026, weekly limits in Claude Code are 25% higher than before the promotion, for Pro, Max, Ultra, Team and seat-based Enterprise (same article).

So the limits are shared, but Anthropic can change Claude Code's weekly allowance without touching the other products. If you read that Claude Code has entirely separate limits from chat, that is wrong as of November 2026.

What this means in practice

• Heavy chat days cost Claude Code time. If you do a lot of long conversations, large file uploads or research-style tasks, expect less left for coding. Anthropic lists conversation length, file size, tools and multi-step tasks among the things that use up limits (best practices article).
• A surprise drop may not be Claude Code. Before you blame a Claude Code session for a sudden jump in usage, remember that phone or browser chats count too. Why Claude Code uses up your limit so fast covers the causes that are in Claude Code itself.
• A model limit is not extra room. On plans that have one, the Fable weekly limit is a share of the same weekly allowance, not an extra pool. See how the Fable weekly limit works.

What is not shared with your plan

Plan limits are for subscription sign-ins. If Claude Code uses an ANTHROPIC_API_KEY instead, it bills your API account per token rather than drawing on your plan (Help Center). That is a separate topic, covered in why Claude Code bills your API key instead of your plan.

Team seats have their own limits per member: "Usage limits can't be transferred or shared between team members. Each seat is for one named individual" (Team plan article). Your personal plan and your work seat are separate accounts with separate limits, even though each is shared across that account's chat and Claude Code.

Seeing the shared total

Settings > Usage on the web, in Claude Desktop and in the mobile apps shows the account-wide figures. The places to look are mapped in where to see your Claude limits.

If you work on a Mac, Headroom is a free menu bar app that shows the same account-wide session and weekly percentages for each Claude Code account signed in on that Mac. Because those percentages are the account totals, they also move when you chat on your phone with the same account. Headroom does not track the Claude desktop, web or mobile apps on their own, only accounts that are signed in to Claude Code on this Mac, and it never sends a prompt, so checking costs nothing against your limits. It reads usage figures that Anthropic does not publish as an official interface, so it can stop working until updated if Anthropic changes them. It is independent of Anthropic and not endorsed by it.