The google-deepmind organization on GitHub hosts a wide range of repositories, primarily utilizing Python, Jupyter Notebook, C++, and C. Notable projects include deepmind-research, which contains code for DeepMind publications, and alphafold, the open-source implementation of AlphaFold 2, showcasing their contributions to AI research and development.
Top languages
Public repositories
mujoco
Multi-Joint dynamics with Contact. A general purpose physics simulator.
deepmind-research
This repository contains implementations and illustrative code to accompany DeepMind publications
alphafold
Open source code for AlphaFold 2.
sonnet
TensorFlow-based neural network library
alphafold3
AlphaFold 3 inference pipeline.
pysc2
StarCraft II Learning Environment
weathernext
No description provided for this repository.
lab
A customisable 3D platform for agent-based AI research
gemma
Gemma open-weight LLM library, from Google DeepMind
open_spiel
OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.
graph_nets
Build Graph Nets in Tensorflow
alphageometry
No description provided for this repository.
dm_control
Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.
mujoco_menagerie
A collection of high-quality models for the MuJoCo physics engine, curated by Google DeepMind.
acme
A library of reinforcement learning components and agents
science-skills
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.
dm-haiku
JAX-based neural network library
alphatensor
No description provided for this repository.
mctx
Monte Carlo tree search in JAX
optax
Optax is a gradient processing and optimization library for JAX.
mujoco_playground
An open-source library for sim and real robot demos on top of MuJoCo.
tapnet
Tracking Any Point (TAP)
alphagenome
This API provides programmatic access to the AlphaGenome model developed by Google DeepMind.
open_x_embodiment
No description provided for this repository.
mathematics_dataset
This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty.
concordia
A library for generative social simulation
mujoco_mpc
Real-time behaviour synthesis with MuJoCo, using Predictive Control
mujoco_warp
GPU-optimized version of the MuJoCo physics simulator, designed for NVIDIA hardware.
rlax
No description provided for this repository.
formal-conjectures
A collection of formalized statements of conjectures in Lean.
android_env
RL research on Android devices.
materials_discovery
No description provided for this repository.
synthid-text
No description provided for this repository.
dramatron
Dramatron uses large language models to generate coherent scripts and screenplays.
funsearch
No description provided for this repository.
chex
No description provided for this repository.
xmanager
A platform for managing machine learning experiments
alphagenome_research
Research code accompanying AlphaGenome
meltingpot
A suite of test scenarios for multi-agent reinforcement learning.
ferminet
An implementation of the Fermionic Neural Network for ab-initio electronic structure calculations
superhuman
No description provided for this repository.
reverb
Reverb is an efficient and easy-to-use data storage and transport system designed for machine learning research
opro
official code for "Large Language Models as Optimizers"
disco_rl
Accompanying code for "Discovering State-of-the-art Reinforcement Algorithms" Nature publication
torax
TORAX: Tokamak transport simulation in JAX
long-form-factuality
Benchmarking long-form factuality in large language models. Original code for our paper "Long-form factuality in large language models".
recurrentgemma
Open weights language model from Google DeepMind, based on Griffin.
limit
On the Theoretical Limitations of Embedding-Based Retrieval
hanabi-learning-environment
hanabi_learning_environment is a research platform for Hanabi experiments.
searchless_chess
Grandmaster-Level Chess Without Search
tips
TIPSv2 (CVPR'26) and TIPS (ICLR'25)
dqn
Lua/Torch implementation of DQN (Nature, 2015)
ai-safety-gridworlds
This is a suite of reinforcement learning environments illustrating various safety properties of intelligent agents.
alphastar
No description provided for this repository.
clrs
No description provided for this repository.
dqn_zoo
DQN Zoo is a collection of reference implementations of reinforcement learning agents developed at DeepMind based on the Deep Q-Network (DQN) agent.
dm_pix
PIX is an image processing library in JAX, for JAX.
representations4d
Foundation models for 4D spatial and temporal vision tasks.
videoprism
Official repository for "VideoPrism: A Foundational Visual Encoder for Video Understanding" (ICML 2024)
physics-IQ-benchmark
Benchmarking physical understanding in generative video models
barkour_robot
Barkour Robot: Agile Quadruped Robots by Google DeepMind
spriteworld
Spriteworld: a flexible, configurable python-based reinforcement learning environment
regress-lm
Library for sequence-to-sequence numeric prediction, applicable to any tokenizable input, and allows pretraining and fine-tuning over multiple tasks.
kfac-jax
Second Order Optimization and Curvature Estimation with K-FAC in JAX.
ai-foundations
No description provided for this repository.
alphaprotein-novo
AlphaProtein Novo
alphaevolve_results
No description provided for this repository.
onetwo
No description provided for this repository.
pg19
No description provided for this repository.
card2code
A code generation dataset for generating the code that implements Hearthstone and Magic The Gathering card effects.
alphaevolve_repository_of_problems
No description provided for this repository.
predictingthepast
No description provided for this repository.
xquad
No description provided for this repository.
language_modeling_is_compression
No description provided for this repository.
jeo
Jeo: Jax model training lib for Earth Observation
PGMax
Loopy belief propagation for factor graphs on discrete variables in JAX
md4
Official Jax Implementation of MD4 Masked Diffusion Models
geeflow
GeeFlow - generate and process large-scale geospatial datasets with Google Earth Engine.
habermas_machine
No description provided for this repository.
forest_typology
Datasets to protect Earth's forests and biodiversity
alphageometry2
AlphaGeometry2 symbolic engine (DDAR) with examples
gemma_penzai
A JAX Research Toolkit for Visualizing, Manipulating, and Understanding Gemma Models with Multi-modal Support based on Penzai.
amplio
Amplio: A Lightweight Agent Harness for Robust and Long-Horizon Runs
boxoban-levels
This repository contains levels for boxoban, a box-pushing puzzle game inspired by Sokoban.
alignet
No description provided for this repository.
cryptids
Some Cryptids and Lean proofs
personality_in_llms
No description provided for this repository.
orbit_sfm
No description provided for this repository.
trecvit
No description provided for this repository.
detectorscope
Build AI text detectors whose decisions are mechanistically grounded and human-auditable (EMNLP 2026)
synthidbio
SynthID Bio is a family of methods developed by Google DeepMind for embedding highly detectable yet function-preserving watermarks directly into AI-generated biological sequences and structures.
phasecoder
PhaseCoder: Microphone Geometry-Agnostic Spatial Audio Understanding for Multimodal LLMs
hybrid_rnns_reward_learning
No description provided for this repository.
formal-putnam-like
Lean formalizations of Putnam-like problems
speech-compass
"Making group conversations more accessible. Real-time speaker localization and directional guidance for mobile speech-to-text. ACM CHI 2025 Best Paper Award."
webquest
No description provided for this repository.
reaf-sim
simulation framework for real robot workflows
relicv2
No description provided for this repository.
gr3en
No description provided for this repository.
unpic
Implementation of unPIC from "How to Spin an Object: First, Get the Shape Right", a CVPRW 2026 paper.
Frequently asked questions
What does google-deepmind build on GitHub?
google-deepmind builds various AI and machine learning tools on GitHub, including repositories like deepmind-research and alphafold, which focus on research implementations and breakthroughs in artificial intelligence.
Which programming languages does google-deepmind use?
The primary programming languages used by google-deepmind include Python, Jupyter Notebook, C++, and C. These languages are prevalent in their projects, enabling robust AI research and development.
Are google-deepmind's repositories public?
Yes, all of google-deepmind's repositories on GitHub are public. This transparency allows researchers and developers to access and utilize their innovative AI tools and implementations.
Is this exposure intended?
Monitor Google DeepMind with RepoGuard and get alerted the moment a new public repository appears.
Monitor this account