Skip navigation
MapAnything V1.1 Release is live! 🚨
0:00 / 1:17

MapAnything V1.1 Release is live! 🚨

Nikhil Keetha

@Nik__V__2.1K followers

Visit on X
720 likesDownloadOriginal post

43,410 views

MapAnything V1.1 Release is live! 🚨 ✅ Improved Checkpoints ✅ Model factory to test & train many models ✅ Profiling ✅ New COLMAP demos & voxelization tooling ✅ WAI format Benchmarking Data Time to update the CVPR subs 😉 Comparisons to DA3 (1.1), Pi3X & more info in 🧵👇

21 Comments

Comments from the original post on X (Twitter)

  • Nikhil Keetha's profile picture

    @Nik__V__

    The key to the new checkpoints is just a better initialization 💡 Splitting DINOv2 ViT-G across both the image encoder and the multi-view transformer!

    14 likes

  • Nikhil Keetha's profile picture

    @Nik__V__

    We expand the Modular MapAnything framework to now support inference & training of many feed-forward 3D models by using simple wrappers 🎁

    20 likes

  • Nikhil Keetha's profile picture

    @Nik__V__

    We profile the V1.1 MapA model against the latest SoTA including DA3 & Pi3. MapA is faster! 💨

    15 likes

  • Nikhil Keetha's profile picture

    @Nik__V__

    We also now include benchmarking comparisons to latest concurrent open source SoTA as of Jan 20th 2026. 🌎 MapAnything is better or competitive to latest models with significantly better metric scale accuracy! 💪 See our 3DV 2026 paper for more details:

    14 likes

  • Nikhil Keetha's profile picture

    @Nik__V__

    To make benchmarking super easy and accessible we release our WAI format data ⏬

    14 likes

  • Nikhil Keetha's profile picture

    @Nik__V__

    New improved COLMAP export & input demos with support for adaptive voxelization! For example, below we go from 13 Million points to 300K by removing redundancy - better downstream uses like GSplat 🤗

    16 likes

  • Nikhil Keetha's profile picture

    @Nik__V__

    Check out our code for more details and to try things out! This marks a freeze on the immediate improvements to the repo. Stay tuned for the next step delta 😉

    27 likes

  • Nikhil Keetha's profile picture

    @Nik__V__

    For more info on MapAnything:

    13 likes

  • Kravec Kustov's profile picture

    @cavalry1854

    Cool work, thanks to FAIR team for updates of their model, easy to miss in constant flow of new releases

    2 likes

  • Everlier's profile picture

    @Everlier

    I'm very curious how it'll behave on the generated videos where perspective is subtly (or often not so subtly) wrong

    1 like

  • Nikhil Keetha's profile picture

    @Nik__V__

    We have tested this and outcome depends on the magnitude of error in consistency across the video generation. One of our results in the paper is on a generated video

    1 like

  • Everlier's profile picture

    @Everlier

    That's really interesting, I think this might be a part of larger reward function for video models to form better world models. Thank you so much for your time writing this reply!

    1 like

  • Nikhil Keetha's profile picture

    @Nik__V__

    Yup! This paper does exactly does this:

    2 likes

  • Everlier's profile picture

    @Everlier

    Haha, sorry for wasting event more of your time on sharing the literature with a layman. Thank you!

  • Chaitany's profile picture

    @Chintu5251

    deep learning for 3d is not my type but i found this Interesting.

    1 like

  • Tyler Gibson's profile picture

    @Gibson61350

    How's it with with 360 images?

    1 like

  • Nikhil Keetha's profile picture

    @Nik__V__

    Our model was not trained with them so the camera calibration would not work. However we find the depth estimation to work. Another way of running on 360s is to use a similar logic to MoGe:

    2 likes

  • David Nordström's profile picture

    @davnords

    Big!

    1 like

  • Kosta Derpanis's profile picture

    @CSProfKGD

    💪

    1 like

  • knaiseris's profile picture

    @knaiseris

    Is this realtime?