Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Cosmos-Predict2.5, the latest version of the Cosmos World Foundation Models (WFMs) family, specialized for simulating and predicting the future state of the world in the form of video.
| Date | Stars |
|---|---|
| 2026-07-24 | 1333 |
| 2026-07-25 | 1333 |
| 2026-07-28 | 1337 |
| 2026-07-30 | 1337 |
| 2026-07-31 | 1338 |
| 2026-08-06 | 1338 |
Today
— stars today
This week
+1 stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.07%/day
> [!IMPORTANT]
> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos)
>
> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Predict2.5, Cosmos 3 delivers significantly stronger world prediction capabilities, producing more accurate, coherent, and physically grounded future-state predictions across a wide range of environments and embodiments.
>
> Beyond improving prediction quality, Cosmos 3 unifies capabilities that previously required multiple specialized models. A single Cosmos 3 model can reason, predict future world states, transfer across domains and modalities, and generate actions and policies for embodied agents within one unified architecture.
>
> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3.
>
> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos
>
> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates.
>
> Thank you for your support of Cosmos-Predict2.5. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities.
<p align="center">
<img src="https://github.com/user-attachments/assets/28f2d612-bbd6-44a3-8795-833d05e9f05f" width="274" alt="NVIDIA Cosmos"/>
</p>
<p align="center">
<a href="https://www.nvidia.com/en-us/ai/cosmos">Product Website</a>  | 🤗 <a href="https://huggingface.co/collections/nvidia/cosmos-predict25-68bb63255f2fc206c5e5b346">Hugging Face</a>  | <a href="https://arxiv.org/abs/2511.00062">Paper</a>  | <a href="https://research.nvidia.com/labs/dir/cosmos-predict2.5">Paper Website</a> | <a href="https://github.com/nvidia-cosmos/cosmos-cookbook">Cosmos Cookbook</a>
</p>
NVIDIA Cosmos™ is a platform purpose-built for physical AI, featuring state-of-the-art generative world foundation models (WFMs), robust guardrails, and an accelerated data processing and curation pipeline. Designed specifically for real-world systems, Cosmos enables developers to rapidly advance physical AI applications such as autonomous vehicles (AVs), robots, and video analytics AI agents.
Cosmos World Foundation Models come in three model types which can all be customized in post-training: [cosmos-predict](https://github.com/nvidia-cosmos/cosmos-predict2.5), [cosmos-transfer](https://github.com/nvidia-cosmos/cosmos-transfer2.5), and [cosmos-reason](https://github.com/nvidia-cosmos/cosmos-reason1).
## News!
* [February 23, 2026] Released Predict2.5 /Robot/Action-Cond Distillation [guide](https://github.com/nvidia-cosmos/cosmos-predict2.5/blob/main/docs/post-training_video2world_action.md#4-distillation), and Predict2.5 /Robot/Policy [models](https://huggingface.co/nvidia/Cosmos-Predict2.5-2B/tree/main/robot/policy) (RoboCasa, Libero), inference and post-training [recipe](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/post_training/predict2/cosmos_policy/post_training.html) in cosmos-cookbook.
* [December 19, 2025] Released Cosmos-Predict2.5-2B Diffusers support via [Hugging Face](https://huggingface.co/docs/diffusers/main/en/api/pipelines/cosmos#diffusers.Cosmos2_5_PredictBasePipeline), Cosmos-Predict2.5-2B Text2World distilled checkpoint on [Hugging Face](https://huggingface.co/nvidia/Cosmos-Predict2.5-2B/tree/main/base/distilled) and Distillation [guide](docs/distillation.md).
* [December 5, 2025] Released Cosmos-Predict2.5-14B [base models](https://huggingface.co/nvidia/Cosmos-Predict2.5-14B), [inference](docs/inference.md) and [post training for DreamGen](docs/post-training_video2world_gr00t.md). Also added the Cosmos-Predict2.5B robot/multiview-agibot [model](https://huggingface.co/nvidia/Cosmos-Predict2.5-2B/tree/main/robot/multiview-agibot), and [inference](docs/inference_robot_multiview-agibot.md).
* [November 25, 2025] Added Blackwell + ARM inference support, along wExcerpt of 13,524 characters
Read on GitHub
10
9
9
3
2
2
2
2
1
Nigel Nelson · United States
1
Miguel Martin · NVIDIA · United States
1
1
1
1
Ming-Yu Liu 劉洺堉 · NVIDIA · United States
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:10e74021a007bc94, topic:video-generation