Anthropic is quietly winning favor with engineers and hobbyists with its tools designed to simplify and automate coding.
Why it matters: Everyone is talking about Claude Code, reinforcing the fact that the AI race is not a binary battle between Google and OpenAI.
The big picture: A mass market, easy-to-use coding tool is a game changer for the 99% of people who, until now, had to rely on other people's software to do anything from building custom data tools to automating repetitive work to building apps or websites.
- And Claude Code — Anthropic's AI coding agent powered by models like Claude Opus 4.5 — is having a moment.
Zoom in: Professionals and dabblers alike say Claude's latest tool is best for coding and outperforms others like Cursor, GitHub Copilot and even Gemini 3 Pro.
- The tool can read an entire codebase, plan complex changes, write and debug code autonomously, run commands and loop for hours on tasks.
Catch up quick: The term vibe coding — describing projects in natural language instead of code — was coined early last year, but most vibe coding tools still required some coding.
- Even as recently as early summer 2025, users still had to understand what a tool was doing in order to use it, Dan Shipper, co-founder and CEO of AI subscription service Every, tells Axios.
- "You still really had to understand the underlying architecture, and maybe you still needed to go look at the code," Shipper says. "It would get lost or go off the rails."
Claude Code changed this by letting users talk directly to an agent and giving Claude full read/write access to files. "You just tell it to do something, and it works," Shipper says, calling it an "infinite vibe coding machine."
State of play: The explosion of online excitement about the tool isn't about one dramatic breakthrough, but a result of different factors coalescing.
- "The underlying models keep getting better," former OpenAI board and executive director at Georgetown's Center for Security and Emerging Technology member Helen Toner tells Axios. "It's not that it's a huge change, but it's a noticeable improvement."
- Another feature working in its favor: Claude Code runs in the terminal, where programmers actually work.
- But Toner thinks the real catalyst was the winter holiday that gave people the extra hours to set up Claude and experiment with it.
Claude Code might be the latest darling, but some say OpenAI's models have also recently reached a kind of escape velocity.
- "It genuinely feels to me like GPT-5.2 and Opus 4.5 in November represent an inflection point — one of those moments where the models get incrementally better in a way that tips across an invisible capability line," developer Simon Willison wrote in his blog late last year.
- "We are in a new era of autonomous coding. You can build astonishingly complex apps without looking at a single line of code," Shipper writes on the Every blog.
Zoom out: Improved vibe coding has the ability to change who gets to make things and whether or not you even need to program.
- "It's also a tremendous change in developer productivity, because once you're at that level, you're not just typing in a command and telling the agent to do something, and then just waiting. You're active."
- Shipper says this can give one engineer the productivity of four or five.
Yes, but: There's a big difference between being able to generate a working prototype via vibe coding and being able to create and iterate secure, enterprise-ready applications.
- It often still takes someone with true programming skills to turn the former into the latter.
What we're watching: The rise of Claude Code and other coding assistants could finally show that AI systems can act as semi-autonomous agents rather than just chatbots that serve up answers.
Editor's note: This story has been updated to reflect Helen Toner's role at Georgetown.
