Over 49,000 developers in 177 countries responded to the 2025 Developer Survey covering 62 questions on 314 technologies and new focus areas like AI agents and large language models. Cargo emerged as the most admired tool in cloud and infrastructure, with Terraform, Homebrew, Make, and APT also highly regarded. Anthropic's Claude Sonnet led admiration among LLMs while OpenAI GPT models were the most desired. AI tools appear in 84% of workflows, with nearly half using them daily, yet trust is limited. AI agents show productivity gains for users despite limited mainstream adoption. Remote and hybrid work remain common and job satisfaction improved slightly.
The 2025 Developer Survey from Stack Overflow gathered input from over 49,000 developers in 177 countries. The survey covered 62 questions on 314 technologies, with new focus areas including AI agent tools, large language models (LLMs), and developer community platforms. Now in its fifteenth year, it provides a broad look at the state of software development worldwide. Stack Overflow 2025 Developer Survey results: Most admired and desired technologies
In cloud development and infrastructure tools, Cargo ranked as the most admired, with 71% of respondents rating it highly. Terraform, Homebrew, Make, and APT also scored well in admiration and desire among developers. Rust's continued rise is tied closely to the popularity of Cargo. Among large language models, Anthropic's Claude Sonnet was the most admired, with OpenAI GPT models as the most desired. Gemini Reasoning and Gemini Flash also appeared prominently in the results.
AI tools remain a significant part of development workflows, with 84% of respondents either using or planning to use them. Daily usage was reported by 47% of developers, and 51% of professional developers use AI tools every day. Despite this, trust remains limited, with only 3% highly trusting AI outputs and 46% expressing some level of distrust. AI agents have yet to see mainstream adoption, with 38% of developers stating they have no plans to use them. However, among those who do, 69% report increased productivity and 70% say agents reduce time on tasks.
Collection
[
|
...
]