Claude Code

TL;DR. Claude Code is Anthropic's agentic coding tool that reads codebases, edits files, runs commands, and handles git workflows through natural language. It runs in the terminal, VS Code, JetBrains IDEs, desktop, and web across macOS, Linux, and Windows. It targets professional developers and engineering teams from startups to enterprises.

Overview

Claude Code is Anthropic's agentic coding tool that reads codebases, edits files, runs commands, and integrates with development workflows through natural language commands. It is available as a terminal CLI, an IDE extension, a desktop application, and via the web, running on macOS, Linux, and Windows. Anthropic describes it as a tool that lets developers delegate routine engineering tasks so they can focus on higher-level design decisions.

Anthropic, PBC was founded on January 26, 2021 in San Francisco, California by seven former OpenAI employees: siblings Dario Amodei (CEO) and Daniela Amodei (President), along with Jared Kaplan, Jack Clark, Chris Olah, Ben Mann, Sam McCandlish, and Tom Brown. The company focuses on building safe, steerable AI systems and is structured as a public benefit corporation. Claude Code launched as a research preview in February 2025 and transitioned to general availability in May 2025. By the time of Anthropic's Series G announcement in February 2026, Claude Code had surpassed $2.5 billion in run-rate revenue, more than doubling since the start of 2026, with weekly active users also doubling since January 2026.

Anthropic has raised over $57 billion in total funding across multiple rounds. The most recent confirmed round is a $30 billion Series G (February 2026) led by GIC and Coatue, valuing the company at $380 billion post-money. Prior rounds include a $13 billion Series F in September 2025 at a $183 billion valuation led by ICONIQ, and a $2 billion round in March 2025 at a $61.5 billion valuation. Key strategic investors include Amazon, Google, Microsoft, and NVIDIA. The company is headquartered at 500 Howard Street, San Francisco, and has grown to approximately 2,500 employees as of 2026, with some third-party estimates ranging higher. Key links: claude.com/product/claude-code, github.com/anthropics/claude-code, code.claude.com/docs/en/overview, anthropic.com/news.

Products & Features

Claude Code is Anthropic's agentic coding tool that operates directly in a developer's terminal, IDE, web browser, and mobile device, handling the full development workflow from codebase onboarding to PR submission. It uses agentic search to map and understand entire repositories, then makes multi-file edits, runs tests, executes shell commands, and integrates with GitHub and GitLab to turn issues into pull requests without context-switching.

Platform availability. Claude Code runs on macOS, Linux, and Windows. It is accessible via the CLI, a native VS Code extension (also compatible with Cursor and other VS Code forks via the Open VSX registry), an official JetBrains plugin supporting IntelliJ IDEA, PyCharm, WebStorm, GoLand, PhpStorm, and Android Studio, Claude Code on the web (a cloud-hosted surface launched October 2025), a desktop app, and mobile push notifications. A GitHub Action (anthropics/claude-code-action) runs the full Claude Code runtime inside GitHub Actions runners for automated PR fixes and issue triage.

Key features. Subagents allow delegating specialized tasks (e.g., backend API work while the main agent builds the frontend) in parallel. Hooks are deterministic scripts that fire at lifecycle points like PreToolUse and PostToolUse for security enforcement, linting, and test runs. Skills are reusable, filesystem-based capabilities stored in .claude/skills/. Plugins bundle skills, subagents, commands, hooks, and MCP server definitions as a single installable unit. Checkpoints automatically save code state before each change and support instant rewind via the /rewind command. Extended thinking lets Claude reason through complex problems before acting. MCP (Model Context Protocol) connectors link Claude Code to external services such as Slack, Linear, databases, and monitoring tools.

Routines. Routines package a prompt, repositories, and connectors into an automated cloud configuration with scheduled (cron), API (HTTP POST trigger), or GitHub event triggers. They run on Anthropic-managed cloud infrastructure and keep working when the developer's laptop is closed. Available on Pro, Max, Team, and Enterprise plans, routines can be created from the web, desktop app, or CLI via /schedule.

Models and API. Claude Code supports multiple models. Claude Opus 4.5 (released November 24, 2025, model ID claude-opus-4-5-20251101) is priced at $5 per million input tokens and $25 per million output tokens, a roughly 66% reduction from Opus 4.1. It offers a 200,000-token context window and up to 64,000-token outputs. The Claude Agent SDK (formerly the Claude Code SDK), available in Python and TypeScript, provides the same tools, agent loop, context management, and permissions framework that power Claude Code for building custom agents in CI/CD pipelines and production applications. Official client SDKs are available in seven languages.

Pricing. Claude Code is included with Pro ($17/month annual, $20 monthly), Max 5x ($100/month), Max 20x ($200/month), Team Premium ($100/seat/month, 5-seat minimum), and Enterprise plans. Team Standard seats at $20/seat do not include Claude Code. Users can also pay per-token via the Anthropic API. A free tier exists for the Claude chat product but does not include Claude Code.

Security & Compliance

Anthropic maintains a strong security and compliance posture for Claude Code, built on enterprise-grade certifications, permission-based architecture, and OS-level sandboxing. The underlying infrastructure holds SOC 2 Type II, ISO 27001:2022, ISO/IEC 42001:2023 (AI Management Systems), and HIPAA-ready configuration with BAA available, as documented on Anthropic's official Privacy Center and Trust Center.

Data Handling and Retention. For commercial API and Enterprise users, Anthropic does not use customer data for model training without express permission. Standard API inputs and outputs are deleted within 30 days. Zero Data Retention (ZDR) is available for the Claude Messages API and for Claude Code when used with Commercial organization API keys or through Claude Enterprise, meaning customer data is not stored at rest after the API response is returned. Consumer plan users (Free, Pro, Max) who opt into model improvement may have data retained in de-identified format for up to 5 years. Data residency options are available, and a Compliance API provides activity feed data retained for 6 years.

Enterprise Security Features. Claude Enterprise supports SAML/OIDC SSO (via WorkOS), SCIM provisioning, role-based access controls, configurable data retention, spend limits, and audit logging. The Compliance API exposes an Activity Feed for monitoring. Audit log exports are metadata-based and do not include chat or project content.

Claude Code-Specific Protections. Claude Code defaults to read-only permissions, requiring explicit approval for file edits, command execution, and network requests. A sandboxed bash tool provides filesystem and network isolation using OS-level primitives (Linux bubblewrap, macOS seatbelt), reducing permission prompts by 84% internally. Write access is confined to the working directory. API keys are stored in macOS Keychain or protected by file permissions on other platforms. Claude Code on the web runs each session in an isolated, Anthropic-managed VM with credential isolation via a secure proxy.

Known Incidents. In March 2025, a Claude Code npm package (v2.1.88) inadvertently included a 59.8 MB source map file, exposing internal TypeScript source code. Anthropic stated no customer data or credentials were exposed and attributed it to a release packaging error. Separately, Check Point researchers disclosed vulnerabilities (CVE-2025-59536) in Claude Code configuration files that could allow remote code execution and API key exfiltration via malicious repositories; these were reported July to October 2025 and patched by Anthropic. Anthropic also detected and disrupted state-sponsored threat actors attempting to misuse Claude for cyber espionage, demonstrating active abuse monitoring.

User Feedback & Adoption

Claude Code has achieved the highest developer satisfaction rating among AI coding tools, with 46% of surveyed developers naming it their "most loved" tool, far ahead of Cursor (19%) and GitHub Copilot (9%). The Pragmatic Engineer AI tooling survey, conducted between January and February 2026 with approximately 1,000 respondents, found Claude Code had overtaken both competitors to become the most-used AI coding tool just eight months after its May 2025 general availability launch. Senior engineering leaders were especially enthusiastic, with the tool twice as popular among directors and senior leaders as at less senior levels.

On G2, Claude Code averages 4.9 out of 5 stars based on 15 reviews, with 100% of reviewers recommending it. Praise centers on deep whole-codebase understanding, terminal and IDE integration (VS Code, JetBrains), MCP and multi-agent capabilities, and the ability to ship robust code quickly. Non-engineers report being able to build real applications with it. The Stack Overflow 2025 Developer Survey (65,000+ respondents) found 61% of developers who used both Claude Code and GitHub Copilot rated Claude Code as more accurate for complex debugging and refactoring, while 73% preferred Copilot for routine code completion.

The dominant complaint across G2, Reddit, Capterra, and Discord is usage limit friction. Developers report token allotments burning through in 10 to 15 minutes of Sonnet usage on Pro plans, with opaque rate-limit thresholds and cooling periods that disrupt workflows. The Register reported in January 2026 that developers observed roughly 60% reductions in effective token limits, which Anthropic attributed to the expiration of a holiday usage bonus rather than a deliberate cut. A March 2026 incident dubbed a "token drain crisis" by DevOps.com saw Max-tier subscribers reporting quota exhaustion in under 20 minutes, prompting a critical GitHub issue (#41930) with widespread community corroboration. Additional complaints include high costs on Max plans, no built-in internet search, context resetting between sessions, and the agent occasionally looping or making overconfident decisions on messy codebases.

Adoption has been exceptionally rapid. Claude Code reached $1 billion in annualized revenue within six months of launch and $2.5 billion by February 2026, according to Anthropic's Series G announcement. Enterprise subscriptions quadrupled since the start of 2026, with enterprise use representing over half of all Claude Code revenue. Deloitte's rollout to approximately 470,000 employees, announced in October 2025, represents the largest enterprise AI deployment on record. By company size, 75% of developers at small companies use Claude Code, while enterprises with 10,000+ employees still favor GitHub Copilot (56%), a divergence attributed to procurement cycles rather than developer preference. SemiAnalysis data indicates Claude Code accounts for approximately 4% of all public GitHub commits as of February 2026, with projections exceeding 20% by year-end.

Monetization & Business Model

Claude Code is monetized through two channels: bundled access via Anthropic's Claude subscription plans (Pro, Max, Team, Enterprise) and pay-as-you-go API billing priced per million tokens. There is no standalone Claude Code subscription; it ships as a feature across every paid tier.

Subscription plans. Pro includes Claude Code at $20/month (or $17/month billed annually at $200 up front). Max offers two tiers, $100/month (5x Pro usage) and $200/month (20x Pro usage). Team plans support 5 to 150 seats at $25/seat/month ($20 annual) for Standard and $125/seat/month ($100 annual) for Premium, with Claude Code available on both seat types and mix-and-match allowed. Enterprise pricing starts at $20/seat with usage cost scaling by model and task, adding SSO, SCIM, audit logs, HIPAA readiness, custom data retention, and admin spend controls. A free tier exists but does not include Claude Code. Usage limits are shared across Claude web/desktop/mobile and Claude Code within a single subscription.

API pricing (per million tokens). Sonnet 5: $2 input / $10 output (introductory rate through August 31, 2026, then $3/$15). Opus 4.8: $5 / $25. Haiku 4.5: $1 / $5. Fable 5: $10 / $50. Prompt caching, batch processing (50% discount), and US-only inference (1.1x pricing) are available. Developers can configure Claude Code to bill API credits when subscription limits are exceeded.

Revenue and trajectory. Claude Code launched in May 2025 and reached approximately $500 million in annualized revenue within three months, per Forbes. By February 2026 it had grown to roughly $2.5 billion in annualized revenue, with business subscriptions quadrupling in the six weeks after January 1, 2026. Anthropic's overall ARR surged from about $1 billion in December 2024 to over $30 billion by April 2026. Over 1,000 enterprise accounts now spend more than $1 million annually on Claude. Anthropic raised $30 billion at a $380 billion valuation in February 2026 and was reportedly seeking a further $50 billion at a valuation above $900 billion in April 2026, with an IPO anticipated as early as October 2026. The company is not yet profitable, with gross margins around 40% and compute spend projected to roughly match revenue in 2026.

Leadership & Team

Anthropic, the company behind Claude Code, was founded in 2021 by seven former OpenAI researchers and engineers. CEO Dario Amodei and President Daniela Amodei (siblings) lead the company, with all seven co-founders still active as of mid-2026.

Board of Directors: Dario Amodei, Daniela Amodei, Yasmin Razavi (Spark Capital), Jay Kreps (Confluent CEO), Reed Hastings (Netflix co-founder), Chris Liddell (former Microsoft CFO, appointed February 2026), and Vas Narasimhan (Novartis CEO, appointed April 2026 via the Long-Term Benefit Trust).

LTBT Trustees: Neil Buddy Shah, Richard Fontaine, and Mariano-Florentino Cuellar.

Company growth: Anthropic had roughly 2,300 employees by end of 2025, more than double its size months earlier. The company raised $65 billion in Series H funding in May 2026 at a $965 billion post-money valuation, led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital.

Culture and values: Anthropic is structured as a Public Benefit Corporation with seven stated values including "act for the global good," "ignite a race to the top on safety," "be helpful, honest, and harmless," and "put the mission first." Notably, all employees share the same title, "Member of Technical Staff," by design, to encourage cross-functional work. Dario Amodei operates with only one direct report (his chief of staff), while Daniela Amodei manages the rest of the workforce. The seven co-founders have pledged 80% of their fortunes to combat AI-driven inequality.

Target Audience & Use Cases

Claude Code targets software developers and engineering teams across individual, SMB, and enterprise segments, with the strongest adoption among professional developers working in medium to large organizations. The tool is a command-line agent that runs in the terminal and supported IDEs (VS Code, Cursor, JetBrains), designed to assist across the entire software development lifecycle from codebase exploration to production maintenance.

Primary personas. Individual developers on Pro or Max plans, and engineering teams on Team or Enterprise plans, form the core audience. Enterprise plans include Claude Code with every seat, while Team plans offer it as a premium seat upgrade. Admins get spend controls, usage analytics (lines of code accepted, suggestion accept rate), managed policy settings, and a Compliance API for regulated industries.

Top use cases (drawn from Anthropic's official use-case guide): (1) Fixing failing tests by tracing code paths and proposing edits; (2) Understanding unfamiliar code during onboarding or inheritance; (3) Refactoring across multiple files with a reviewable plan before edits; (4) Writing tests that match existing project conventions; (5) Reviewing and summarizing pull requests with full repository context. Additional workflows include CI/CD automation, large-scale refactors, issue resolution end-to-end via MCP integrations, and turning recurring tasks into reusable Skills.

Industry verticals and customer examples. Official Anthropic customer stories span telecommunications (Cox Communications), software (Vercel, Box, Cursor, Warp, Replit), life sciences (Garvan Institute), professional services (Quantium, Blank Metal), and cybersecurity (Kai). Anthropic also cites Behavox (compliance/security, hundreds of developers) and Altana (supply chain AI/ML, 2-10x development velocity acceleration) as early Claude Code enterprise adopters. The Government of Alberta uses Claude Code for cybersecurity vulnerability review across government systems.

Company size sweet spot. Adoption ranges from startups (Delight.ai, Juno, Warp) to large enterprises (Cox Communications, Box, Vercel). The Team plan targets single departments or small companies, while Enterprise adds SAML/SCIM, audit logs, HIPAA/BAA compliance, and data residency for organizations scaling company-wide.

Geographic focus. Customer stories span North America, Europe, and Asia Pacific, with US-only inference available as an Enterprise option for data residency requirements.

Tags & Categories

Primary Category: AI Coding

Categories: AI Coding, AI DevOps

Tags: agentic coding, cli coding tool, terminal ai, multi-file editing, codebase understanding, git workflows, github integration, gitlab integration, ci/cd automation, vs code extension, jetbrains plugin, developer productivity, anthropic, claude

Claude Code is an agentic coding tool by Anthropic that lives primarily in the terminal and extends to IDE plugins (VS Code, JetBrains), a desktop app, and a web interface. It is classified as an AI Coding tool, with secondary relevance to AI DevOps given its CI/CD, GitHub Actions, and GitLab CI/CD integrations for automating PR reviews and issue triage. The tool understands entire codebases through agentic search, makes multi-file edits, executes shell commands, and handles git workflows via natural language commands. It also supports Slack, Chrome, and mobile (iOS) integrations, and exposes a Claude Agent SDK for building custom agentic workflows. Core use cases include code onboarding, turning issues into PRs, debugging, feature building, and automated code review.

Impact & Recommendations

Claude Code has become the fastest-scaling software product in history, reaching $1 billion in annualized run-rate revenue within six months of its May 2025 public launch and $2.5 billion by February 2026, according to Anthropic's own Series G announcement. That trajectory outpaced even ChatGPT's early growth and positioned Anthropic as the enterprise AI coding leader, with the broader company hitting a $30 billion run rate by April 2026.

Market position and competitive landscape. Claude Code competes in a three-way race with GitHub Copilot and Cursor. Per the Pragmatic Engineer 2026 survey of 15,000 developers, Claude Code was rated "most loved" by 46% of respondents and used as the primary tool by 71% of developers who regularly use AI agents. It dominates at small and mid-size companies (75% adoption), while GitHub Copilot retains the enterprise lead at large organizations (56% adoption among 10,000+ employee firms). An estimated 4% of all public GitHub commits worldwide are now authored by Claude Code, per SemiAnalysis. Anthropic's terminal-native, agentic architecture (direct file system access, persistent state, multi-day workflows) is its primary differentiator versus Copilot's IDE-extension model and Cursor's repository-aware editor.

Key growth signals. Weekly active users doubled between January and February 2026. Business subscriptions quadrupled since the start of 2026, with enterprise use now representing over half of Claude Code revenue. The average developer spends 20 hours per week with the tool. Claude Opus 4.5 achieved 80.9% on SWE-bench Verified, leading the benchmark field.

Risk factors. A weeks-long performance decline in March-April 2026, traced to three engineering missteps, triggered significant user backlash and subscription cancellations. Cybersecurity firm TrustedSec measured a 47% drop in code quality during the period. Compute constraints are an ongoing structural risk: Anthropic acknowledged infrastructure was "stretched" at peak hours, and an OpenAI internal memo claimed Anthropic was "operating on a meaningfully smaller" compute curve. Security researchers at Check Point disclosed vulnerabilities (including CVE-2026-21852) that could expose API keys. Shift to consumption-based enterprise pricing may triple costs for heavy users.

ICP fit. Claude Code is best suited for engineering teams tackling complex, multi-file refactoring, architecture decisions, and large-scale debugging. It is less optimal for simple autocomplete workflows where Copilot excels, and teams with strict compliance perimeters or existing GitHub Enterprise contracts face slower procurement cycles.

Outlook. The category is expanding beyond coding into broader knowledge work via Cowork, Anthropic's plugin-based agent platform. With an IPO reportedly under consideration for late 2026 and continued model advancements, Claude Code's trajectory remains strong, though compute capacity and quality consistency will determine whether it holds its lead against fast-moving competitors.

Chiri Analysis

Chiri Score: 88/100

Verdict

Best for:

• Engineering teams doing complex, multi-file refactoring, architecture work, and large-scale debugging

• Professional developers at startups and mid-size companies who want a terminal-native agentic workflow

• Enterprises needing SSO, SCIM, audit logs, spend controls, and a Compliance API with Claude Code on every seat

• Teams automating CI/CD, PR reviews, and issue triage via GitHub Actions and GitLab CI/CD

• Developers who want whole-codebase understanding without manually selecting context files

Not for:

• Developers who primarily want fast, low-cost routine autocomplete, where GitHub Copilot is preferred by 73% of dual users

• Heavy users sensitive to usage-limit friction and opaque rate-limit thresholds or consumption-based cost spikes

• Large organizations with existing GitHub Enterprise contracts and strict procurement perimeters facing slow adoption cycles

• Teams requiring built-in internet search or fully consistent quality, given documented performance-decline incidents

• Users needing a free tier, since Claude Code is not included in Anthropic's free plan

Head-to-head

Frequently Asked Questions

How much does Claude Code cost?

Claude Code is included with Anthropic's Pro plan at $20/month ($17/month billed annually, $200 up front), Max at $100/month (5x usage) or $200/month (20x usage), Team Premium at $125/seat/month ($100 annual), and Enterprise starting at $20/seat with usage-based scaling. Team Standard seats at $25/month do not include it. There is no standalone Claude Code subscription; developers can also pay per token via the Anthropic API.

Is Claude Code SOC 2 compliant?

Anthropic's infrastructure holds SOC 2 Type II and ISO 27001 certifications, with reports available at the Anthropic Trust Center. Documentation also references ISO/IEC 42001:2023 and HIPAA-ready configuration with a BAA available, though some certification details are noted as unverified. Zero Data Retention is available for Claude Code via commercial organization API keys or Claude Enterprise.

Who are Claude Code's main competitors?

Claude Code competes directly with GitHub Copilot and Cursor. In the Pragmatic Engineer 2026 survey, 46% of developers named Claude Code their most loved AI coding tool versus 19% for Cursor and 9% for GitHub Copilot. Copilot retains an enterprise lead at 10,000+ employee firms; Cursor competes as a repository-aware editor.

Is Claude Code good for enterprise?

Yes. Enterprise plans include Claude Code with every seat and add SAML/OIDC SSO, SCIM provisioning, role-based access controls, audit logging, spend controls, usage analytics, and a Compliance API. Enterprise use represents over half of Claude Code revenue, and Deloitte deployed it to roughly 470,000 employees in October 2025.

What platforms and IDEs does Claude Code support?

Claude Code runs on macOS, Linux, and Windows via terminal CLI, a native VS Code extension (also compatible with Cursor and other VS Code forks via Open VSX), an official JetBrains plugin (IntelliJ IDEA, PyCharm, WebStorm, GoLand, PhpStorm, Android Studio), a desktop app, and the web. It integrates with GitHub, GitLab, Slack, and Chrome.

Who makes Claude Code?

Claude Code is built by Anthropic, a public benefit corporation founded January 26, 2021 in San Francisco by seven former OpenAI employees, including CEO Dario Amodei and President Daniela Amodei. Boris Cherny, a former Meta Principal Engineer and author of Programming TypeScript, is the creator and Head of Claude Code.

How fast is Claude Code growing?

Claude Code launched to general availability in May 2025 and reached $1 billion in annualized run-rate revenue within six months and $2.5 billion by February 2026, more than doubling since the start of 2026. Weekly active users also doubled since January 2026. It became the most-used AI coding tool within eight months of launch.

What is Claude Code best used for?

Claude Code excels at complex, multi-file refactoring, whole-codebase understanding, large-scale debugging, test fixing, and turning GitHub or GitLab issues into pull requests. In the Stack Overflow 2025 survey, 61% of dual users rated it more accurate than Copilot for complex debugging and refactoring, while 73% preferred Copilot for routine code completion.

---

Reviewed by Chiri Atlas Research Desk (AI Tooling Analyst) on 2026-07-07.