JetBrains extends Google Cloud partnership to bring Gemini models to AI assistant

Software development tools provider JetBrains has teamed up with Google Cloud to bring the cloud giant’s Gemini models to its AI Assistant coding tool.

Building on the pair's existing relationship, the collaboration will see JetBrains leverage Google Cloud’s Vertex AI to expand the tool’s existing pool of large language models (LLMs) in a bid to drive AI-powered developer productivity.

The addition means JetBrains’ AI Assistant is the first offering of its kind to combine functionality from OpenAI’s GPT-4o, Gemini, as well as several of the company’s own proprietary models.

In an announcement, Kirill Skrygan, CEO of JetBrains, said the rapid advancement of generative AI meant the company had to find partners that could “innovate and create the most capable models.”

“JetBrains has a long and rich history of partnering with Google Cloud across many areas,” he commented. “We had an exclusive preview of Gemini Pro and Ultra during our AI Hackathon with Google Cloud a few months ago and quickly saw their potential.”

JetBrains’ suite of dedicated integrated development environments (IDEs) for multiple programming languages integrates with the firm’s AI Assistant to generate suggestions and general function and efficient code.

Available inside the IDE, the AI Assistant can generate code for requests, suggest fixes for specific issues, or refactor a function. Developers can also use the chat to gain answers to specific questions, as well as generate tests and documentation.

JetBrains said its AI Assistant is now the only tool that selects the most optimal LLM for each specific task to enable developers to spend more time on creative challenges and minimize repetitive tasks.

The addition of Gemini Pro 1.5 and Flash 1.5 on Google Cloud’s Vertex AI will now expand the offering’s use cases thanks to its long context window, advanced reasoning, as well as high performance levels. 

Specifically, JetBrains said Gemini 1.5 will be of particular use to use cases in which cost efficiency at high volume and low latency are key requirements.

The functionality is expected to roll out “in the coming weeks,” the firm added, at which point enterprise users will be able to configure exactly which LLM their team members can use.

The collaboration is the latest move from JetBrains in its mission to expand its AI offering. Earlier this year, the firm released full line code autocompletion powered by locally run AI models to help ensure that data processing and analysis occur directly on the developer’s device for faster offline code completion and minimal latency.

“We are excited to have our collaboration with JetBrains expanding, now with the Gemini model family being supported in JetBrains AI,” commented Warren Barkley, Google Cloud’s senior director of product management for Vertex AI.

“We have enjoyed working closely with talented developers from JetBrains and seeing some amazing new AI-driven capabilities they've built leveraging unique capabilities of Gemini models, such as 1M long context window and advanced reasoning.

“We are looking forward to these powerful innovations to be put in the hands of millions of developers using JetBrains products globally.”