Claude
by Anthropic
Anthropic's AI assistant: chat, Artifacts, Projects, Research, and agentic Cowork
Last reviewed 2026-06-20
Claude is Anthropic's consumer and enterprise AI assistant, available on the web (claude.ai), macOS and Windows desktop apps, and iOS and Android. It is the assistant brand that sits alongside Anthropic's developer API and the separate Claude Code coding agent. The core experience is conversational: it answers questions, writes and edits text and code, analyzes uploaded files, searches the web with citations, and builds interactive Artifacts (tools, visualizations, and small apps). Projects keep persistent context per topic, Memory carries continuity across sessions, and Connectors (built on Anthropic's open Model Context Protocol) wire Claude into Gmail, Google Workspace, Slack, Jira, and Zapier. Most of Claude's surface is assistant- or copilot-grade. A newer, narrower slice is genuinely agentic: Research plans and runs multiple chained web and Workspace searches into a cited report, and Cowork (a research-preview desktop agent) reads, edits, and creates files inside a folder you choose and can browse the web to complete multi-step tasks. Those agentic modes still run under human direction and review, so the representative experience is a copilot rather than an autonomous agent. It is powered by Anthropic's own Claude model family.
What it can do
Chat, write, and analyze files
AssistantAnswers questions, drafts and edits text and code, and analyzes uploaded documents and images on request, with web search and inline citations.
sourceBuild interactive Artifacts
AssistantTurns a prompt into shareable tools, visualizations, and small apps that persist across sessions and can call Claude's API to become interactive micro-applications.
sourceRun multi-step Research
SupervisedOperates agentically, running multiple chained searches across the web and connected Google Workspace (Gmail, Calendar, Docs) and returning a cited report in minutes; the human sets the question and reviews the output.
sourceWork on local files with Cowork
SupervisedA research-preview desktop agent that, within a folder you choose, reads, edits, and creates files and can browse the web to organize folders, build spreadsheets, or prepare reports.
sourceConnect to external tools via MCP Connectors
CopilotConnectors built on Anthropic's open Model Context Protocol link Claude to Gmail, Google Workspace, Slack, Jira, and Zapier so it can read context and trigger actions in those tools.
source
Strengths
- +Strong writing, reasoning, and coding plus interactive Artifacts in one assistant
- +Open MCP Connectors wire it into Gmail, Workspace, Slack, Jira, and Zapier
- +Generous free tier across web, desktop, and mobile; Research and Cowork add real agentic work
Limitations
- −Most of the product is assistant/copilot grade; the agentic modes are newer and narrower
- −Cowork ships as a research preview and (for now) on the desktop app
- −Locked to Anthropic's own Claude models (no model choice)
Overview
Claude is Anthropic's AI assistant, available on the web at claude.ai, on macOS and Windows desktop apps, and on iOS and Android. It is the assistant brand that sits alongside Anthropic's developer API and the separate Claude Code coding agent, and it runs on Anthropic's own Claude model family.
What it does
The core experience is conversational: Claude answers questions, drafts and edits text and code, analyzes uploaded files and images, and searches the web with citations. Artifacts turn a prompt into shareable tools, visualizations, and small apps that persist across sessions. Projects keep persistent context per topic and Memory carries continuity across conversations. A newer agentic slice goes further: Research operates agentically, running chained searches across the web and connected Google Workspace into a cited report, and Cowork (a research-preview desktop agent) reads, edits, and creates files inside a folder you choose and can browse the web to finish multi-step tasks. Those modes run under human direction and review.
Integrations & setup
Connectors, built on Anthropic's open Model Context Protocol (MCP), link Claude to Gmail, Google Workspace, Slack, Jira, Zapier, and Microsoft 365. Anthropic created MCP as an open standard, and Claude also supports voice mode and Skills (specialized capability modules). The developer API exposes the same models for building on top.
Pricing
Freemium. A free tier covers web, desktop, and mobile with daily limits. Pro is $20/month ($17/month billed annually) and adds Research, Claude Code, Cowork, unlimited Projects, and more usage. Max is $100/month (5x usage) or $200/month (20x usage). Team is $25/seat/month ($20 annually), and Enterprise is custom (seat price plus usage at API rates) with SSO, audit logs, and a HIPAA-ready option.
Best for / not for
Best for individuals and teams who want a strong general assistant for writing, reasoning, coding, and research, with optional agentic Research and Cowork on top. Less suited to buyers who need a mature end-to-end autonomous agent today or who want a choice of underlying models.
Alternatives
Google Gemini and Microsoft Copilot are the direct general-assistant rivals with deep ecosystem integration; Perplexity competes on cited research; ChatGPT Agent overlaps on agentic task execution. For Anthropic's coding-specific agent, see Claude Code.
What people are saying
We aggregate real LinkedIn discussion into sentiment for the agents people search most. Claude isn't tracked yet, want it added? Request tracking.
FAQ
Is Claude an autonomous agent?+
Mostly no. The chat, Artifacts, and Projects experience is assistant- and copilot-grade. Research is a supervised agent that plans and runs chained searches into a cited report, and Cowork is a research-preview desktop agent that works on local files. Both run under human direction and review, so the representative experience is a copilot.
How is Claude different from Claude Code?+
Claude is Anthropic's general AI assistant for chat, writing, research, and file work across web, desktop, and mobile. Claude Code is a separate terminal-native agentic coding tool that edits codebases, runs commands, and handles git. Both are powered by Anthropic's Claude models and a Pro subscription includes access to both.
What does Claude cost?+
There is a free tier, Pro at $20/month (or $17/month billed annually), and Max at $100/month (5x usage) or $200/month (20x usage). Team is $25/seat/month ($20 annually) with premium seats at $125/$100, and Enterprise is custom (seat price plus usage at API rates).
Sources
- Claude product overview (Anthropic / claude.com) · accessed 2026-06-20
- Claude pricing (claude.com) · accessed 2026-06-20
- Introducing Claude's Research (claude.com blog) · accessed 2026-06-20
- Anthropic launches Cowork, a Claude Desktop agent (VentureBeat) · accessed 2026-06-20
- Anthropic (Wikipedia, company background) · accessed 2026-06-20
Last reviewed 2026-06-20