Google Opens Frontier AI Models to 200 European Scientists

Google DeepMind and Google Cloud are giving European scientists access to some of their most advanced AI models, along with the computing power needed to put them to work. Announced on 14 October 2026, the initiative will provide up to 200 researchers with direct access to Google's frontier AI models and $3 million in cloud computing credits. 

The tools cover everything from predicting extreme weather to studying DNA and understanding how molecules interact. The commitment is part of Google's support for the European Union's AI for Science programme, which is working to make advanced AI technology more accessible to researchers across the region.

For scientists who need powerful computers to process large amounts of information, getting access to the right technology can be a challenge in itself. Google's latest commitment is intended to help address that problem, giving selected researchers more of the resources they need to carry out their work.

Google DeepMind and Google Cloud Join Europe's RAISE Initiative

Google's new commitment will support the Resource for AI Science in Europe (RAISE) initiative, an EU programme launched in November 2025 to help scientists make greater use of artificial intelligence in their research.

RAISE works as a virtual research institute, bringing together computing resources, scientific data, funding and expertise from across Europe. Researchers can draw on these shared resources without everything needing to be housed in one physical institution.

The European Commission introduced the programme as part of its wider European AI in Science Strategy, which aims to strengthen the region's scientific research capabilities and support the development and use of advanced AI. The RAISE pilot is backed by €107 million in funding through Horizon Europe, the EU's research and innovation programme.

Google's contribution adds specialist AI technology and cloud resources to that existing effort. Its research division, DeepMind, develops the models, while Google Cloud provides the computing infrastructure that researchers can use to run demanding workloads.

In its announcement, Google said the commitment is intended to help address the computing gap faced by European scientists and speed up scientific discovery. It also forms part of the company's efforts to support Europe's position in advanced scientific research.

Five AI Models Covering Research From DNA to Extreme Weather

The models included in the programme aren't all designed for the same kind of research. Google is making five specialist tools available, each developed to help scientists investigate different problems.

For researchers working in biology, AlphaFold 3 predicts the three-dimensional structures of molecules and how they interact. This can help scientists understand how proteins, DNA and other molecules behave together, including interactions that are important when developing new medicines.

There's also AlphaGenome, which looks more closely at genetic information. It predicts how changes in DNA can affect the processes that control genes, giving scientists another way to investigate how genetic variations influence biological functions and disease.

Weather forecasting is another area covered by the programme. Google's WeatherNext family of models uses AI to predict weather conditions, including the development and movement of tropical cyclones. DeepMind has already worked with the US National Hurricane Center on using the technology to support forecasts of dangerous storms.

For scientists studying the environment, AlphaEarth Foundations combines information from satellites and other Earth observation sources to build detailed representations of the planet's land and coastal waters. Researchers can use this information to investigate changes in forests, agricultural land and other environments over time.

The fifth tool, AlphaEvolve, takes a different approach. It's an AI coding agent powered by Google's Gemini models, designed to find and improve algorithms, the instructions computers use to solve problems. DeepMind has previously used it for mathematical research and to improve computing processes, including work involving its own infrastructure.

Google Cloud Adds Computing Credits to the Programme

Alongside the models, Google Cloud's contribution will give participating scientists access to computing resources for their research.

This is an important part of the announcement because scientific AI can involve processing enormous amounts of information. Researchers working with genetic sequences, weather data or satellite observations may need substantial computing power to run their experiments and analyse the results.

Google says the combination of its models and cloud resources is intended to help European researchers tackle these challenges more quickly. The company also hopes the initiative will strengthen Europe's scientific competitiveness by giving research teams access to advanced technology.

For now, the announcement confirms the scale of Google's commitment and the tools involved, but leaves some practical questions unanswered. Google hasn't specified when researchers will receive access, how the computing credits will be allocated or how long the arrangements will last.

Final Thoughts: Google Extends Its Scientific AI Support to Europe

Google's announcement adds a new European research commitment to its existing work on making scientific AI tools available to researchers.

The company has previously announced similar partnerships with the UK government and the US Department of Energy, offering access to specialist models for scientific work. Its latest commitment brings that approach to Europe's RAISE programme, where the focus will be on helping selected researchers use advanced AI to investigate problems across several scientific fields.