ZenkenAI
Published: Updated:

Claude Usage Limits Explained: Free, Pro, Max — All Plans (2026 Edition)

Claude
Table of contents 68 items
Zenken is featured as an OpenAI customer story

Short version: Claude is not truly unlimited. Every plan has some kind of cap, and the differences between plans are large. This article uses Anthropic’s official help center and changelogs to give you the picture as of April 2026.

The latest flagship — Claude Opus 4.7 — released April 16, 2026, is now available on Pro and above.

📊 Claude usage limits — quick reference (April 2026)

PlanMonthlyMessagesClaude CodeCoworkContext
Free$0Dynamic❌❌200K
Pro$20~45 / 5 hrs40–80 hrs/wk✅1M
Max 5x$100~225 / 5 hrs140–280 hrs/wk✅1M
Max 20x$200~900 / 5 hrs240–480 hrs/wk✅1M
Team Standard$25/seatPer-seat✅✅1M
Team Premium$125/seat ($100/yr)Per-seat (6.25×)✅✅1M
EnterpriseContact salesCustom✅✅1M

Model access by plan (April 2026)

ModelFreeProMaxTeam / Enterprise
Claude Opus 4.7❌✅✅✅
Claude Opus 4.6❌✅✅✅
Claude Sonnet 4.6✅✅✅✅
Claude Haiku 4.5✅✅✅✅
Claude Code❌✅✅✅
Cowork❌✅✅✅

Reset timing

  • 5-hour limits: rolling window (5 hours from your first message)
  • Daily limits: 00:00 UTC
  • Weekly limits: 7 days from your first message, at 00:00 UTC

Claude Opus 4.7 (April 2026)

Anthropic released Claude Opus 4.7 on April 16, 2026 — the latest and most capable generally available model.

Highlights

  • Major software-engineering gains: clear improvements over Opus 4.6 on hard coding tasks
  • Better instruction following: more reliable on detailed requirements and format specs
  • 3× vision performance: image long-edge up to ~2,576 px (~3.75 megapixels) — large improvement on dense charts and chemistry-style notation
  • Stronger self-verification: on long tasks, Claude verifies its own output before delivering, producing more consistent results
  • New xhigh effort level: finer control over the reasoning vs. speed tradeoff
  • Task Budgets (public beta): cap usage on a per-task basis on the Claude platform
  • /ultrareview command: a dedicated code-review session in Claude Code

Benchmarks

  • Agentic coding, tool use, computer use, financial analysis: surpasses GPT-5.4 and Gemini 3.1 Pro — the top-rated generally available model again
  • Terminal-Bench / SWE-bench: continues Opus 4.6’s industry-leading performance
  • Complex, long-running tasks: early testers report it as “comfortable to delegate the hardest engineering work to”

API pricing

API pricing matches Opus 4.6: $5 input / $25 output per 1M tokens. Model ID: claude-opus-4-7. Note that the updated tokenizer can produce 1.0×–1.35× more tokens for the same input — measure before scaling.

Model release timeline

ModelReleasedNotable for
Claude Opus 4.7April 16, 2026Latest flagship. Adaptive Thinking only — no Extended Thinking.
Claude Sonnet 4.6February 17, 2026Big jumps on coding, knowledge work, agentic planning
Claude Opus 4.6February 5, 20261M context, Adaptive Thinking, Agent Teams
Claude Opus 4.5November 24, 2025effort parameter introduced
Claude Sonnet 4.5September 29, 2025Coding / agent specialization
Claude Opus 4.1August 5, 2025Stronger agent tasks
Claude Opus 4 / Sonnet 4May 22, 2025Generation-4 base models

Free plan limits

Free uses Claude Sonnet 4.6, with limits.

Message limits

  • Short messages: ~40 messages/day
  • Long messages or messages with attachments: ~20–30 messages/day
  • Reset: every 5 hours

February 2026 Free-plan expansion

Effective February 11, 2026, Free added:

  • File creation: Word, Excel, PowerPoint, PDF
  • Third-party connectors: Slack, Canva, Notion, Figma, etc.
  • Skills: save and reuse workflows
  • Conversation compaction: automatically summarize long conversations to stay within context

Other restrictions

  • Claude Opus 4.7 / 4.6: unavailable
  • Claude Code: unavailable
  • Cowork: unavailable
  • MCP Apps: basic features available
  • Projects: limited
  • File upload: yes (within limits)
  • Web search: yes

Caveats

As of April 2026, Free limits are dynamic — they vary based on server load and query complexity rather than a fixed daily quota.

When you hit the limit, you wait until the next reset (5 hours). During peak hours, Free users are deprioritized — responses get slower or you hit the cap sooner.

Claude Pro ($20/month)

The most popular paid plan, designed for individual users.

Message limits (April 2026)

  • Messages: ~45 messages per 5 hours (~216/day)
  • Claude Code: 40–80 hours/week of Sonnet 4.6
  • Context window: 1M tokens (with Opus 4.7 / 4.6 / Sonnet 4.6)

How limits work

When you hit the cap, you wait until the next reset (5 hours). The 5-hour window is rolling — it starts from your first message.

How many messages you can send depends on:

  • Message length
  • Conversation length (accumulated context)
  • Whether attachments are included

What Pro adds

  • Claude Opus 4.7 access: latest flagship (Opus 4.6 still selectable)
  • Claude Code: AI coding agent
  • Cowork: autonomous desktop tasks
  • MCP Apps: third-party service integrations
  • Projects: file creation and management
  • Google Workspace: Gmail, Docs integrations
  • Research: deep multi-step research
  • Priority access: stable access during peak hours
  • Extra usage: buy more at API rates when capped

Claude Max ($100–$200/month)

For power users. Introduced May 1, 2025.

Max 5x ($100/month)

  • Messages: ~225 / 5 hours (5× Pro)
  • Claude Code: 140–280 hrs/week Sonnet 4.6, 15–35 hrs/week Opus 4.7 / 4.6
  • Auto-switch: at 20% Opus usage, downshifts to Sonnet 4.6
  • Memory: conversation history retained
  • Priority: maintained during high-traffic periods

Max 20x ($200/month)

  • Messages: ~900 / 5 hours (20× Pro)
  • Claude Code: 240–480 hrs/week Sonnet 4.6, 24–40 hrs/week Opus 4.7 / 4.6
  • Auto-switch: at 50% Opus usage, downshifts to Sonnet 4.6
  • Excel integration: deeper Excel features
  • Extra usage: buy more at API rates if capped

Max plan features

  • Cowork: autonomous desktop work
  • Agent Teams: multiple agents in parallel (research preview)
  • Claude in PowerPoint: integration with PowerPoint

Pro / Max usage is shared across Claude (chat), Claude Code (coding), and Cowork (desktop agent). Whatever tool you use, it draws from the same counter.

Claude Team ($25–$125/seat/month)

Optimized for organizations. Pricing structure refreshed January 2026.

Pricing

  • Standard seat: $25/seat/month
  • Premium seat: $125/seat/month (monthly) / $100/seat/month (annual)
  • 5–150 seats

Standard vs. Premium

ItemStandardPremium
Monthly$25/seat$100–125/seat
UsagePer-seat limitsPer-seat (6.25× Pro)
Claude Code✅✅
Cowork✅✅
Claude Opus 4.7✅✅

Team highlights

  • Per-seat limits: each member has individual limits (was previously a shared pool)
  • SSO: integration with corporate identity providers
  • Enterprise search: search internal data
  • Microsoft 365 integration: Office product integration
  • Admin console: user management and usage dashboards

Claude Enterprise

Custom plans for larger organizations.

Highlights

  • Extended context window: 1M tokens (Opus 4.7 / 4.6 / Sonnet 4.6)
  • Custom usage: configured to your needs
  • Role-based access: permission management
  • Audit logs: compliance support
  • SCIM: automated user provisioning
  • SLA: contractual uptime
  • Dedicated support: priority channels

Limits

Most Enterprise users won’t notice usage limits day to day, but you do operate under terms of service.

Claude Code limits

Claude Code is available on Pro, Max, Team (Standard / Premium), and Enterprise.

Plan-by-plan Claude Code limits

PlanSonnet 4.6Opus 4.7 / 4.6
Pro40–80 hrs/weekLimited
Max 5x140–280 hrs/week15–35 hrs/week
Max 20x240–480 hrs/week24–40 hrs/week

Claude Code 2.1 (January 2026)

Released January 7, 2026:

  • Agent Teams (research preview): multiple agents working in parallel on different tasks
  • Skills improvements: hot reload — skill changes apply in real time

Shared usage

Claude (chat), Claude Code, and Cowork share a usage counter. Heavy use of Claude Code reduces what’s left for chat and Cowork.

Important: weekly limits added (August 28, 2025)

On August 28, 2025, Anthropic introduced weekly limits in response to abuse by some power users:

  • Weekly limits supplement the existing 5-hour rolling window
  • Max users can buy extra usage at API rates if they exceed it
  • Per Anthropic, “most users won’t notice” — fewer than 5% of users are affected at current usage patterns

Claude Cowork

Cowork was announced as a research preview on January 12, 2026 and moved to general availability (GA) on April 9, 2026. It’s a desktop-based AI agent — sometimes described as “Claude Code for non-engineers”.

Available on

  • Pro: ✅
  • Max: ✅
  • Team Premium: ✅
  • Enterprise: ✅
  • Team Standard: ✅
  • Free: ❌

Platform support

  • macOS: January 12, 2026
  • Windows: February 10, 2026

Capabilities

  • Local file operations: read, write, create files on your PC
  • Multi-step task autonomy: completes work that requires multiple steps
  • Plugin support (since January 30, 2026): connect to third-party tools and services

Note

Cowork uses more tokens than chat — it plans, executes, and verifies each step. Claude / Claude Code / Cowork share a counter, so heavy Cowork use reduces what’s left for the others.

Why limits exist

Three main reasons:

  1. Server resource management: powerful models like Opus 4.7 require enormous compute. Without limits, a few power users could starve everyone else.

  2. Fair access: with millions of users worldwide, resources have to be shared.

  3. Cost control: each Opus 4.7 response is expensive to generate. Without caps, operating costs would balloon.

How to check your limit

As of April 2026, the Claude app does not show a real-time “remaining messages” dashboard.

What you can see

  • Approaching-limit notifications
  • Reset time when you hit the cap

What you can’t see

  • Current remaining messages
  • Detailed usage history

What happens when you hit the limit

Free

  • Notification appears
  • Wait 5 hours for reset
  • Upgrade prompt

Pro / Max

  • Notification appears
  • Wait for the 5-hour or weekly reset
  • Pro / Max can buy extra usage at API rates
  • Upgrade is also an option

Team / Enterprise

  • Per-seat limits per member
  • Premium seats get 6.25× Pro usage
  • Temporary restrictions only on detected abuse

What is a context window?

The context window is the amount of information Claude can “read” in a single session.

Per plan

Plan / modelContext windowApprox. equivalent
Free (Sonnet 4.6 / Haiku 4.5)200K tokens~150,000 words (~300 pages)
Pro / Max / Team / Enterprise (Opus 4.7 / 4.6 / Sonnet 4.6)1M tokens~750,000 words (~1,500 pages)

Context Compaction

Introduced with Opus 4.6 (beta), Context Compaction summarizes and compresses long conversations automatically, letting them continue while preserving the important parts.

What happens if you exceed the context window?

  • Some of the content you provided isn’t considered
  • The earliest parts of the conversation get “forgotten”
  • The big picture across very long files becomes harder to track

This matters most when uploading multiple large files for analysis.

Major updates 2025–2026

May 2025: Max plan and Claude 4

May 1, 2025: Anthropic introduced the Max plan. May 22: Claude Opus 4 and Sonnet 4 released.

  • Max plan: $100/month (5x) and $200/month (20x) tiers
  • Claude Opus 4: state of the art on coding, 72.5% on SWE-bench
  • Claude Sonnet 4: significant upgrade from Sonnet 3.7

August 2025: weekly limits + Opus 4.1

  • August 5: Claude Opus 4.1 (stronger agent tasks)
  • August 28: weekly limits introduced for power-user abuse

September 2025: Claude Sonnet 4.5

September 29, 2025:

  • Specialized for coding and agents
  • State of the art on SWE-bench Verified
  • 30+ hour multi-step task capability

November 2025: Claude Opus 4.5

November 24, 2025:

  • The most powerful model at the time
  • effort parameter (token-usage control)
  • Practical pricing for production use

January 2026: Cowork, Claude Code 2.1, MCP Apps

A big month for the ecosystem:

  • January 7: Claude Code 2.1 (Agent Teams, Skills improvements)
  • January 12: Cowork announced (macOS research preview)
  • January 22: New Constitution (AI behavior principles) published
  • January 26: MCP Apps (third-party app integration)

February 2026: Opus 4.6 and major Free expansion

  • February 5: Claude Opus 4.6 (1M context beta, Adaptive Thinking, Agent Teams)
  • February 10: Cowork on Windows
  • February 11: Free plan adds file creation, third-party connectors, Skills
  • February 17: Claude Sonnet 4.6 (big jumps on coding, knowledge work, agent planning)

March 2026: 1M context window GA

  • March 13: 1M-token context window moves to GA — standard on Opus 4.6 / Sonnet 4.6, automatic on Max / Team / Enterprise, available on Pro

April 2026: Cowork GA and Opus 4.7

  • April 8–9: Claude Managed Agents public beta — deploy and manage agents via API
  • April 9: Cowork moves from research preview to GA
  • April 16: Claude Opus 4.7 — major gains on software engineering, vision, and instruction following; back at the top of the public-model leaderboards. Also introduces /ultrareview for Claude Code, the xhigh effort level, and Task Budgets (beta)

Summary: which plan should you pick?

Free works for you if you…

  • Use Claude lightly (a few times a day)
  • Are still trying it out
  • Can’t justify a budget yet
  • Are happy with Sonnet 4.6
  • Want file creation and connectors (now free as of February 2026)

Claude Pro is right if you…

  • Use it daily for work or study
  • Need access to Claude Opus 4.7 (the latest flagship)
  • Want Claude Code or Cowork
  • Best value-for-money ⭐⭐⭐⭐⭐

Claude Max is right if you…

  • Are a specialist or developer
  • Use Claude all day
  • Use Claude Code, Cowork, Agent Teams heavily
  • Need priority access during peak hours

Claude Team / Enterprise is right for organizations that…

  • Have multiple users
  • Need data security guarantees
  • Need SSO, audit logs, etc.
  • Need an SLA or dedicated support

Picking the right plan

  1. Frequency: how many messages per day
  2. Model: do you need Opus 4.7 / 4.6, or is Sonnet 4.6 enough?
  3. Claude Code / Cowork: do you need the agent products?
  4. Budget: $0–$200/month range

Most individual users land on Claude Pro ($20/month). ~45 messages per 5 hours and a 1M context window cover most daily business work.

For pricing details and payment see Claude Pricing Guide. New to Claude? See Claude — A Beginner’s Guide.

For Claude Code-specific pricing see Claude Code Pricing, and for the all-plans feature comparison see Claude Plan Comparison.

Staying current

Claude limits change often. Check these for the latest:

Last updated: April 13, 2026

Note: limits may change without notice. Verify the current state at Anthropic’s official sources.

References