These are the most popular AI coding tools among engineers
AI coding tools are gaining traction across the tech industry. A recent survey sheds light on which services are most popular among engineers.
In May, Jellyfish, which helps companies manage developer teams, surveyed 645 full-time professionals in various engineering roles, including individual contributors, managers, and executives. Respondents came from companies ranging from small teams with fewer than 10 people to enterprises with more than 500 engineers.
The survey findings shed new light on the explosive growth and impact of AI coding tools in software development.
Jellyfish found that 90% of engineering teams are now using AI in their workflows, up from 61% just one year ago. Almost a third have formally supported and widely adopted AI tools, while another 39% are actively experimenting with them. Only 3% of respondents reported no AI usage and no plans to change that.
Crucially, 48% of respondents reported using two or more AI coding tools, suggesting teams are taking a diversified, exploratory approach by evaluating multiple solutions simultaneously rather than standardizing on a single platform.
The leaders
The leader among AI coding tools was GitHub Copilot from Microsoft, with 42% of surveyed engineers naming it their tool of choice, according to the survey. Google's Gemini Code Assist was second, while Amazon Q (formerly CodeWhisperer) and Cursor were tied at third.
These four tools formed the dominant tier of AI-powered code assistance platforms, but there were several other services in the mix, too, according to the report.
The study explicitly excluded general-purpose generative AI tools like ChatGPT to focus on products designed specifically for software engineering. This distinction highlights the growing specialization of AI solutions tailored to the needs of development teams.
According to the report, 62% of engineers said they've achieved at least a 25% boost in velocity and productivity thanks to AI coding tools, and 8% reported a doubling of their output. Less than 1% believe AI is slowing them down.
Human-AI hybrid workflows
Looking ahead, 81% of respondents believe that at least a quarter of today's engineering work will be automated by AI within the next five years. Yet, the trend isn't toward full automation; it's toward collaboration.
"While AI can help creatives, AI itself is not creative," one engineering leader put it in their survey response.
"If you have smart people using AI who also understand the topic/issue they are going after, magic happens," the person added. "Otherwise, you have people that desperately just want to look like they have done something amazing, but don't really understand the issues they have just created with the help of AI."
With productivity gains already measurable and adoption rising, the current crop of AI coding tools, led by GitHub Copilot, Gemini, Amazon Q, and Cursor, appears to be setting the foundation for a hybrid future where software engineers and AI systems co-create the next generation of digital products.