Google DeepMind ORG

AlphaFold 3 inference pipeline.

Multi-Joint dynamics with Contact. A general purpose physics simulator.

MuJoCo is a fast physics engine by Google DeepMind for simulating robots, skeletons, and jointed bodies, widely used in robotics research and machine learning training.

Open source code for AlphaFold 2.

DeepMind's open-source inference code for AlphaFold 2, which predicts the 3D structure of a protein from its amino-acid sequence using deep learning.

A library of reinforcement learning components and agents

JAX-based neural network library

This repository contains implementations and illustrative code to accompany DeepMind publications

Monorepo of code, environments, and datasets that accompany DeepMind research papers, with one sub-folder per project covering RL, physics, language, and biology work.

TensorFlow-based neural network library

StarCraft II Learning Environment

A Python library from DeepMind that turns StarCraft II into an AI research environment, letting you train reinforcement learning agents to play the game and benchmark them on focused mini-game scenarios.

A customisable 3D platform for agent-based AI research

A customizable 3D game-world from Google DeepMind for training AI agents using reinforcement learning, where agents navigate mazes and solve puzzles to earn rewards inside a Quake III-based simulation.

No description.

Gemma open-weight LLM library, from Google DeepMind

Build Graph Nets in Tensorflow

OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.

No description.

Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.

A collection of high-quality models for the MuJoCo physics engine, curated by Google DeepMind.

No description.

Monte Carlo tree search in JAX

A TensorFlow implementation of the Differentiable Neural Computer.

Optax is a gradient processing and optimization library for JAX.

No description.

An open-source library for GPU-accelerated robot learning and sim-to-real transfer.

GDM Science Skills to speed up agentic scientific workflows with better grounding and higher token efficiency. Integrate insights from AlphaGenome, AFDB, UniProt and 30+ other databases and tools.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty.

This API provides programmatic access to the AlphaGenome model developed by Google DeepMind.

Tracking Any Point (TAP)

A JAX research toolkit for building, editing, and visualizing neural networks.

No description.

Convolutional neural network model for video classification trained on the Kinetics dataset.

Real-time behaviour synthesis with MuJoCo, using Predictive Control

bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent

A library for generative social simulation

A Graph Neural Network Library in Jax

No description.

GPU-optimized version of the MuJoCo physics simulator, designed for NVIDIA hardware.

RL research on Android devices.

No description.

Dramatron uses large language models to generate coherent scripts and screenplays.

No description.

tree is a library for working with nested data structures

A collection of formalized statements of conjectures in Lean.

No description.

No description.

A platform for managing machine learning experiments

A suite of test scenarios for multi-agent reinforcement learning.

An implementation of the Fermionic Neural Network for ab-initio electronic structure calculations

Reverb is an efficient and easy-to-use data storage and transport system designed for machine learning research

Research code accompanying AlphaGenome

No description.

official code for "Large Language Models as Optimizers"

No description.

Accompanying code for "Discovering State-of-the-art Reinforcement Algorithms" Nature publication

TORAX: Tokamak transport simulation in JAX

Open weights language model from Google DeepMind, based on Griffin.

Dermatology ddx dataset, Jax implementations of Monte Carlo conformal prediction, plausibility regions and statistical annotation aggregation from our recent work on uncertain ground truth (TMLR'23 and ArXiv pre-print).

hanabi_learning_environment is a research platform for Hanabi experiments.

On the Theoretical Limitations of Embedding-Based Retrieval

No description.

Grandmaster-Level Chess Without Search

No description.

Lua/Torch implementation of DQN (Nature, 2015)

This is a suite of reinforcement learning environments illustrating various safety properties of intelligent agents.

No description.

Restoring and attributing ancient texts using deep neural networks

No description.

No description.

TIPSv2 (CVPR'26) and TIPS (ICLR'25)

Dataset to assess the disentanglement properties of unsupervised learning methods

No description.

Minimal and scalable research codebase in JAX, designed for rapid iteration on frontier research in LLM and other autoregressive models.

DQN Zoo is a collection of reference implementations of reinforcement learning agents developed at DeepMind based on the Deep Q-Network (DQN) agent.

An interactive HTML pretty-printer for machine learning research in IPython notebooks.

Libraries, tools and tasks created and used at DeepMind Robotics.

No description.

Official repository for "VideoPrism: A Foundational Visual Encoder for Video Understanding" (ICML 2024)

Barkour Robot: Agile Quadruped Robots by Google DeepMind

Library for sequence-to-sequence numeric prediction, applicable to any tokenizable input, and allows pretraining and fine-tuning over multiple tasks.

A algebraic word problem dataset, with multiple choice questions annotated with rationales.

No description.

Second Order Optimization and Curvature Estimation with K-FAC in JAX.

No description.

Implementation of Meta-Learning with Latent Embedding Optimization

Benchmarking physical understanding in generative video models

Multi-object image datasets with ground-truth segmentation masks and generative factors.

No description.

No description.

No description.

No description.

No description.

A code generation dataset for generating the code that implements Hearthstone and Magic The Gathering card effects.

Lean math proofs generated by AlphaProof Nexus and accompanying natural language prose proofs.

No description.

No description.

No description.

No description.

Algorithms for Privacy-Preserving Machine Learning in JAX

Gemini 2025 ICPC World Finals Code Submissions

Implementation of Spectral Inference Networks, ICLR 2019

No description.

Loopy belief propagation for factor graphs on discrete variables in JAX

No description.

Jeo: Jax model training lib for Earth Observation

Official Jax Implementation of MD4 Masked Diffusion Models

No description.

This is the 0.4 release of the Arcade Learning Environment (ALE), a platform designed for AI research. ALE is based on Stella, an Atari 2600 VCS emulator.

This repository contains the 3D shapes dataset, used in Kim, Hyunjik and Mnih, Andriy. "Disentangling by Factorising." In Proceedings of the 35th International Conference on Machine Learning (ICML). 2018. to assess the disentanglement properties of unsupervised learning methods.

No description.

Neural network verification in JAX

Transformer Grammars: Augmenting Transformer Language Models with Syntactic Inductive Biases at Scale, TACL (2022)

No description.

GeeFlow - generate and process large-scale geospatial datasets with Google Earth Engine.

No description.

No description.

A tool for recording RL trajectories.

No description.

A networking protocol for agent-environment communication

Datasets to protect Earth's forests and biodiversity

No description.

No description.

A JAX Research Toolkit for Visualizing, Manipulating, and Understanding Gemma Models with Multi-modal Support based on Penzai.

This repository contains the code of the distribution shift framework presented in A Fine-Grained Analysis on Distribution Shift (Wiles et al., 2022).

This repository contains levels for boxoban, a box-pushing puzzle game inspired by Sokoban.

ML-guided enzyme engineering

AlphaGeometry2 symbolic engine (DDAR) with examples

No description.

Multi-framework implementation of Deep Kernel Shaping and Tailored Activation Transformations, which are methods that modify neural network models (and their initializations) to make them easier to train.

Code release for the paper, "Proactive Agents for Text-to-Image Generation under Uncertainty"

This repository contains an entailment dataset for propositional logic, and code for generating that dataset. It also contains code for parsing the dataset in python.

No description.

XTR: Rethinking the Role of Token Retrieval in Multi-Vector Retrieval

No description.

No description.

Dataset for the spaceship task from "Metacontrol for Adaptive Imagination-Based Optimization"

No description.

No description.

No description.

No description.

Cluttered MNIST Dataset

No description.

A JAX implementation of Broaden Your Views for Self-Supervised Video Learning, or BraVe for short.

Cephes Mathematical Functions library wrapped for Torch

No description.

No description.

No description.

Unsupervised Data Generated for GeoQuery and SAIL Datasets

No description.

No description.

GenAI evaluation framework, optimized for 100x lower cost 🚀.

A library of surrogate transport models for tokamak fusion.

No description.

Datasets used to train Spatial Language Integrating Model (SLIM) in the ‘Encoding Spatial Relations from Natural Language’ paper.

No description.

Open source code for EigenGame.

A Python library for defining structured command-line flags.

No description.

Real-time visualisation

No description.

No description.

No description.

Simulation Streams is a programming paradigm designed to efficiently control and leverage Large Language Models (LLMs) for complex, dynamic simulations and agentic workflows.

No description.

A Lua package to detect reading of undeclared variables and creating of global variables.

No description.

Torch test module

Gemini evaluations

Code & data for "Towards flexible perception with visual memory" (ICML 2025)

No description.

No description.

No description.

Scripts to help with Torch package documentation

No description.

No description.

LuaJIT wrapper for PLplot

A pretty print library for torch and lua.

No description.

No description.

No description.

No description.

No description.

Latex package for visualizing 2x2 games

No description.

A self-contained JAX implementation of DIScriminator DisAgreement INtrinsic Reward (DISDAIN).

No description.

No description.

No description.

No description.

No description.

No description.

No description.

No description.

"Making group conversations more accessible. Real-time speaker localization and directional guidance for mobile speech-to-text. ACM CHI 2025 Best Paper Award."

Style guide for Lua code.

No description.

No description.

No description.

No description.

No description.

No description.

A Python library for easy access and manipulation of nested data structures.

No description.