author: @kotauchisunsun
Release date: 2026/07/03
We have been operating a site called AI Coding.Info since July 2025.
This is a site that observes usage trends related to AI Coding Agents such as Claude Code, Codex CLI, Github Copilot, and Gemini from a fixed point from information in Github repositories. To determine the use of AI Coding Agent, we conduct daily surveys under the following conditions.
https://ai-coding.info/reports/articles/20260603
AI Coding Agent repository utilization rate was 12.1%, an increase of 1.1% from 11.0% last time.


AI Coding Agent usage rate on July 1, 2026
https://ai-coding.info/?date=2026-07-01#adoption-rate
AI Coding Agent usage rate on June 1, 2026
https://ai-coding.info/?date=2026-06-01#adoption-rate
Trends in AI Coding Agent usage rate from June 1, 2026 to July 1, 2026
https://ai-coding.info/?date=2026-07-01&since=2026-06-01&until=2026-07-01#share-trend
The share by product is as follows.
| Ranking | Product name | Share rate |
|---|---|---|
| 1st place | Codex CLI | 39.1% |
| 2nd place | Claude Code | 32.0% |
| 3rd place | Copilot Agent | 15.9% |
| 4th place | Gemini CLI | 5.7% |
| 5th place | Cursor | 4.7% |

Well, this month's topic is the regulation of Claude Code's latest AI agents such as Mythos and Fable 5, and how will this affect them? That's what I thought. However, it actually decreased slightly. That is a fact. There has been a lot of talk about Fable 5 being regulated in the US, so I don't think the impact has been accurately reflected in the market yet. Therefore, although it is not possible to make an accurate evaluation, the Codex has increased by 2.2%. I think that is a big fact. ClaudeCode is down 0.8% and Copilot Agent is down 0.9%, indicating that switching is progressing. There is also a feeling. This time, Cursor will be acquired by SpaceX. There was news.
https://www.nikkei.com/article/DGXZQOGN16BC70W6A610C2000000/
However, in reality, the usage rate of Cursor has not changed much, with a slight decrease of 0.4%. This is also the case, but since SpaceX's listing has not been very long, the impact will not be felt until a little later. There is a possibility.
AI Coding Agent share rate on 2026/7/1
https://ai-coding.info/?date=2026-07-01&since=2026-06-01&until=2026-07-01#agent-share
AI Coding Agent share rate on June 1, 2026
https://ai-coding.info/?date=2026-06-01&since=2026-06-01&until=2026-07-01#agent-share
The programming language in which AI Coding Agent is used the most is "TypeScript," the second is "Rust," the third is "Python," the fourth is "Go," and the fifth is "C#."

A diagram summarizes the transition of programming languages used by AI Coding Agent in 2026.

What you can see from this is that TypeScript continues to consistently maintain its number one position. I write a report every month, and I get the impression that the rankings for Python'', Go'', and ``Rust'' are not stable. As shown in the graph earlier, these languages are languages that are constantly changing because the number of repositories they use is close to each other. You could say that. On the other hand, I wonder if "C#" in 5th place is starting to stabilize. That's the impression I get. He has always maintained 5th place except for the February summary, so he seems to be in a position like 1st among the 3 teams.
AI Coding Agent ranking by programming language on July 1, 2026
https://ai-coding.info/?date=2026-07-01&since=2026-06-01&until=2026-07-01#language-agent-rank
AI Coding Agent ranking by programming language on June 1, 2026
https://ai-coding.info/?date=2026-06-01&since=2026-06-01&until=2026-07-01#language-agent-rank
The number of repositories was 1,611 as of June 11, 2026, but it has increased to 1,794 as of July 1, 2026, and the number of repositories using AI Coding Agent has increased by 183.

Changes in the number of repositories using AI Coding Agent from 2026/6/1 to 2026/7/1
https://ai-coding.info/?date=2026-07-01&since=2026-06-01&until=2026-07-01#time-based-bar-chart

https://ai-coding.info/?since=2025-07-01&until=2026-07-01#ai-usage-line-chart
The usage rate on July 1, 2025, when the service started was 2.0%. The utilization rate on July 1, 2026 is 12.1%.
**AI Coding Aent usage rate increased by 10.1% in 1 year. **
Regarding this, what is the usage rate on the first day of each month and how much has the usage rate increased compared to the previous month? I created a graph showing the difference.

Looking at this from an overview, the usage rate of AI Coding Agent increased by approximately 0.5% in 2025 (although it specifically increased by 0.9% on August 1, 2025). Since January 2026, the usage rate has been increasing by more than 1% every month, except for January. therefore,
AI Coding Agent penetration is increasing twice as fast as last year
This is an astonishing result. Why has it progressed so far? I previously discussed this in an article from February.
https://ai-coding.info/reports/articles/20260301
One reason is that the number of Codex users has increased, and AI Coding is now being used in programming languages that were not used before. That was my conclusion. If you think about other reasons, is it financial results? budget? Things like that just crossed my mind. What I mean by this is that in Japan, the fiscal year ends in March, and the new fiscal year often starts in April. However, this is not the case overseas, where the fiscal year starts in January and ends in December. When you think about that, it will be dangerous if you don't do AI Coding in 2026! So, I started properly budgeting from 2026. I thought that such a way of thinking might actually exist. In that case, there is a compromise, and the increase in usage rate on January 1, 2026 is about 0.5%, which is the same as last year. That is actually the usage rate between 2025/12/01 and 2025/12/31, so nothing will change. However, since I started using it from 2026/01/01, the usage rate will start to increase rapidly from 2026/02/01. That's why it permeated in a sort of trickle-down manner. I wonder if there is an aspect like that. I also thought so.
It's been a year since I started a service called AI Coding.Info. Service management is difficult. I can't stop making batches every day, or I'm running out of batches. However, thanks to you, we are able to obtain data for 365 days every day. I started writing the article on July 1st, but on the first anniversary of the service, a bug occurred and I was unable to deploy it. Then, make an emergency correction to the system and release it. Even after a year, bugs that we thought would never occur still occur. I thought deeply. The system now has over 60 million records, making it a fairly large service for an individual to own. The daily aggregation processing on the system was putting a strain on it, making deployment impossible. Since this is the current situation, we need to modify the architecture and system configuration. That's how I feel.