Claude Code Now Available on the Web

Anthropic has officially launched a web-based version of its popular Claude Code AI coding assistant, allowing developers to create and manage multiple AI coding agents directly from their browser. This new web application is currently available to subscribers of the company’s Pro plan at $20 per month, as well as the higher-tier Max plans priced at $100 and $200 monthly. Existing Pro and Max users can access Claude Code on the web by visiting claude.ai and selecting the “Code” tab, the same platform that hosts Anthropic’s consumer chatbot.

This strategic move represents Anthropic’s effort to expand Claude Code beyond its original command-line interface, transforming it from a terminal-exclusive tool into a more versatile web application. The goal is to enable developers to initiate AI coding agents across various environments, increasing flexibility and accessibility. The market for AI-assisted coding has become intensely competitive, with numerous tech giants and startups introducing their own solutions. While Microsoft’s GitHub Copilot was an early leader, newcomers like Cursor, Google, OpenAI, and Anthropic have developed highly capable alternatives, many already accessible via web interfaces. Despite this crowded field, Claude Code stands out as one of the most widely adopted options. Since its broader release in May, user numbers have surged tenfold, and the product now contributes over $500 million in annualized revenue for the company.

Cat Wu, a Product Manager at Anthropic, explained that much of Claude Code’s success stems from the company’s AI models, which have gained significant popularity among developers. Wu also mentioned that the development team intentionally incorporates elements of enjoyment into the product experience. Looking ahead, Anthropic plans to make Claude Code available across more platforms, though the terminal version will likely remain the core environment. Wu emphasized that ensuring the CLI product offers the most intelligent and customizable way to utilize coding agents remains a priority, while simultaneously expanding accessibility through web and mobile interfaces.

Interestingly, Anthropic reports that approximately 90% of the Claude Code product was written using the company’s own AI models. Wu, who previously worked as an engineer, noted that she now rarely writes code manually, instead focusing on reviewing outputs generated by Claude Code. This reflects a broader evolution in AI coding tools. Early versions functioned primarily as autocomplete systems, suggesting code snippets as developers typed. The current generation, including Claude Code, enables the creation of autonomous agents that handle more complex coding tasks independently. This shift is transforming the role of software engineers, who increasingly act as managers overseeing AI assistants rather than writing every line of code themselves.

Not all developers have embraced this change enthusiastically. Recent studies indicate that some engineers actually work slower when using AI coding tools like Cursor. Researchers observed that significant time was spent crafting prompts and waiting for AI responses, potentially reducing productivity. Additionally, these tools sometimes struggle with large, complex codebases, leading engineers to spend extra time correcting inaccurate AI-generated code. Despite these challenges, companies like Anthropic continue to advance their AI coding agent technology. CEO Dario Amodei recently predicted that AI will soon be responsible for writing 90% of code for software engineers. While this may already be happening within Anthropic, widespread adoption across the broader economy might take considerably longer to materialize.

(Source: TechCrunch)