# Python repositories of the `google-deepmind` GitHub organization

Reference catalog of every public Python repository owned by the [google-deepmind](https://github.com/google-deepmind) GitHub organization, captured for the aphrody project.

- **Total Python repos:** 236  (205 active / 31 archived)
- **Snapshot:** 2026-05-21 (GitHub Search API `org:google-deepmind language:python`)
- **Exhaustive table:** [`all-repos.md`](../../../reference/ai/google-deepmind/all-repos.md) — all 236 repos, sorted by stars.

## Thematic pages

Every repo is assigned to exactly one thematic page (the five counts below sum to 236).

| Category | Repos | Page | Focus |
|----------|------:|------|-------|
| Science, datasets & tools | 76 | [catalog/science-datasets-and-tools.md](../../../reference/ai/google-deepmind/catalog/science-datasets-and-tools.md) | Computational biology/chemistry, NLP & reasoning datasets, medical AI, privacy & fairness, causal/Bayesian methods, ML tooling, NAS, research-paper code |
| Reinforcement learning & environments | 52 | [catalog/reinforcement-learning-and-environments.md](../../../reference/ai/google-deepmind/catalog/reinforcement-learning-and-environments.md) | Acme & agent frameworks, MuJoCo physics simulation, multi-agent & gridworld environments, evaluation benchmarks, meta-learning |
| Neural network libraries | 46 | [catalog/neural-network-libraries.md](../../../reference/ai/google-deepmind/catalog/neural-network-libraries.md) | Sonnet (TensorFlow), sequence & language models, graph nets, verification & robustness, vision/media, representation learning |
| JAX ecosystem | 38 | [catalog/jax-ecosystem.md](../../../reference/ai/google-deepmind/catalog/jax-ecosystem.md) | Core NN libraries, optimization & training utilities, testing/verification, graph learning, probabilistic & generative models, scientific simulation |
| Foundation models & AlphaX | 24 | [catalog/foundation-models-and-alphax.md](../../../reference/ai/google-deepmind/catalog/foundation-models-and-alphax.md) | AlphaFold protein structure, mathematical & algorithmic reasoning, the Gemma family, weather/Earth-system modelling, optimisation-based discovery |

## Top 12 by stars

| # | Repo | Stars | Status | Description |
|--:|------|------:|--------|-------------|
| 1 | [alphafold](https://github.com/google-deepmind/alphafold) | 14605 | active | Open source code for AlphaFold 2. |
| 2 | [sonnet](https://github.com/google-deepmind/sonnet) | 9918 | active | TensorFlow-based neural network library |
| 3 | [pysc2](https://github.com/google-deepmind/pysc2) | 8292 | active | StarCraft II Learning Environment |
| 4 | [alphafold3](https://github.com/google-deepmind/alphafold3) | 8054 | active | AlphaFold 3 inference pipeline. |
| 5 | [graphcast](https://github.com/google-deepmind/graphcast) | 6661 | active | GraphCast weather forecasting models |
| 6 | [graph_nets](https://github.com/google-deepmind/graph_nets) | 5401 | active | Build Graph Nets in Tensorflow |
| 7 | [gemma](https://github.com/google-deepmind/gemma) | 5239 | active | Gemma open-weight LLM library, from Google DeepMind |
| 8 | [alphageometry](https://github.com/google-deepmind/alphageometry) | 4841 | active | Olympiad-level geometry theorem solver |
| 9 | [dm_control](https://github.com/google-deepmind/dm_control) | 4586 | active | Software stack for physics-based simulation and RL environments, using MuJoCo. |
| 10 | [learning-to-learn](https://github.com/google-deepmind/learning-to-learn) | 4070 | active | Learning to Learn in TensorFlow |
| 11 | [acme](https://github.com/google-deepmind/acme) | 3986 | active | A library of reinforcement learning components and agents |
| 12 | [mujoco_menagerie](https://github.com/google-deepmind/mujoco_menagerie) | 3485 | active | High-quality models for the MuJoCo physics engine, curated by Google DeepMind. |

See [`all-repos.md`](../../../reference/ai/google-deepmind/all-repos.md) for the complete, exhaustive list of all 236 repositories.

---
_Generated from `docs/reference/ai/data/google-deepmind-repos.json`. Thematic pages under `catalog/` are enriched with real GitHub topics and README excerpts._