Why Claude Code Suddenly Went Viral in June 2025
This article was created manually except for AI-assisted typo detection.
I’m Oikon. I work as an engineer at a foreign IT company.
This article is a polished version of my posts on 𝕏 (formerly Twitter) about why Claude Code became a hot topic.
I’m compiling this as a memo to look back on fondly in the future.
What is Claude Code?
Most people who are interested in reading articles about Claude Code probably already know what Claude Code currently is, so I’ll focus a bit on the past.
- February 24, 2025: Claude Code Pre-release (Research Preview version)
- April 10, 2025: Claude Max Plan launched (Claude Code not yet included)
- May 1, 2025: Claude Code becomes available as unlimited usage with Claude Max Plan
- May 22, 2025: Claude 4 (Opus/Sonnet) announced
Claude Code was released as a Research Preview in February, making it available to the general public.
At that time, it used Claude 3.7, and aside from being a CLI tool, its basic features weren’t much different from those in June 2025.
Even at that point, it had appealing features like:
- Understanding the entire codebase
- More autonomous operation than other AI tools
- Claude 3.7’s excellent coding capabilities
And it had a dedicated following among some users.
I’ve personally relied on it quite a bit for my hobby personal development projects.
After Claude Code was included in the Claude Max Plan from May 1st, you could use it without worrying about API billing.
At that time, OpenAI’s o3 and Google’s Gemini 2.5 Pro were making waves in the AI world, so not many people seemed to be paying attention to Claude Code.
Then things changed dramatically with Anthropic’s announcement on May 22nd.
You can see what was announced in the YouTube video below:
In this article, I’d like to share my personal thoughts on why Claude Code caught fire after May 22nd.
Why Claude Code Went Viral
① Claude 4 (Opus/Sonnet) Performance is Excellent
Claude 3.7 until mid-May already had quite high coding capabilities.
However, competitors OpenAI o3 and Gemini 2.5 Pro were released between March and April, and it was hard to deny that Anthropic had fallen behind in release speed.
Then Claude 4 Opus/Sonnet was released.
If this had flopped, Anthropic would have fallen significantly behind in the AI development race, but Claude 4 performed beyond many people’s expectations.
As of June 2025, it’s probably not an exaggeration to say it has the strongest coding capabilities.
The significant advancement of Claude 4 created a huge response, especially among engineers.
② Unlimited Claude Code Usage with Claude Max Plan
Claude Code itself was already included in the Claude Max Plan’s unlimited usage from May.
This wasn’t widely known, and because Claude 3.7’s performance inevitably looked inferior compared to other generative AI models, there seemed to be few users.
After the Claude 4 series was released, it gained recognition for its performance, and attention also gathered on the unlimited Claude Code usage.
Until then, development with AI editors like Cursor and Windsurf was mainstream, but many people couldn’t go all-in on development because they were worried about API pay-per-use billing.
With the Claude Max Plan, you can use Claude Code for a flat rate, allowing you to rapidly iterate with AI-powered Try & Error without worrying about API costs.
This unlimited usage of Claude Code was a major difference from previous AI services and became a factor that drew attention.
③ IDE Integration with VSCode, Cursor, etc.
Claude Code’s IDE integration was also a major factor that drew attention.
Claude Code was originally pre-released as a CLI tool, and it was already excellent at that point.
However, being a CLI tool, it didn’t seem to resonate with many engineers.
After the May 22nd announcement, IDE integration became available for:
- VSCode (including forks like Cursor and Windsurf)
- JetBrains IDE
Since most current engineers develop with these IDEs, being able to integrate Claude Code into their existing editors dramatically expanded the user base.
When I posted about IDE integration before, the response was significant.
④ More Autonomous AI Operation Than Ever
A characteristic of Claude Code is that when you give it a task, it creates its own TODO list and executes it autonomously.
Claude Code’s strength is being able to proceed with tasks while understanding the entire project.
What this changes compared to conventional AI models is that instead of coding through dialogue with AI, it becomes a collaborator you can delegate work to.
I think it’s getting close to the feeling of outsourcing work to AI.
By clearly defining requirements and preparing tests as guardrails, you can delegate the actual implementation to AI.
Of course, there are still mistakes, but Claude Code is showing the potential of autonomous AI operation.
You could say that developers have shifted from hands-on coding work to upstream tasks of requirements definition and design.
Additional Factor A: Claude Code GitHub Actions
Claude Code GitHub Actions was announced together with the May 22nd announcement.
If you mention @claude in Issues or Pull Requests on GitHub’s UI and ask for tasks, it will complete the tasks entirely within GitHub.
This also became a topic, but it’s pay-per-use billing regardless of the Claude Max Plan.
I made this an additional factor because I’m personally skeptical about whether it directly contributed to Claude Code’s excitement.
Additional Factor B: GitHub Copilot
There was also an interesting quote comment I’d like to share:
Personally, the introduction of GitHub Copilot’s premium request limit system from June is also significant.
GitHub Copilot was (in my understanding) popular among light users as an AI coding tool.
The opinion that users flowed from Copilot to Claude because Copilot was essentially introducing restrictions from June was interesting.
Update (20250609): Pro Plan Unlimited Access
There was an event I forgot to mention.
On June 5, 2025, Claude Code was made available for unlimited use on the Pro Plan as well.
This allowed people to try Claude Code with the Pro Plan ($20) instead of just the Max Plan ($100+), further expanding the user base.
As a complete aside, the phrase “Max or Pro” existed in Claude Code v1.0.7 on June 2nd (it was quickly removed), so I think they originally planned to open it up to the Pro Plan as well.
Summary: The Age of Claude Has Arrived
Personally, I feel Claude Code is a tool that has advanced AI development to the next stage.
It’s unclear whether OpenAI or Google will enter this coding domain in the future, but Claude Code + Claude 4 was undoubtedly an epoch-making event.
I believe the Claude Code experience is a preview of AI development six months to a year from now.
As an Anthropic fan, I’ll continue to watch Claude’s developments closely.
Follow Me on 𝕏!
I also share information on 𝕏, so I’d really appreciate it if you followed me!
References
- (20250225) Claude 3.7 Sonnet and Claude Code
- (20250410) Introducing the Max Plan
- (20250523) Introducing Claude 4