Skip to content
@moonmath-ai

MoonMath.ai

performance for physical AI

MoonMath ai (7)

MoonMath.ai builds the performance layer for Physical AI.

We are a small team of mathematicians & engineers building production-grade acceleration for the next wave of AI systems via low level algorithms and system engineering.

MoonLite

MoonLite is MoonMath’s category of inference acceleration kernels designed for large generative models.

  • LiteAttention: Transforming Video Diffusion with Temporal Sparse Attention
  • LiteFFN: Replaces standard FFN layers with a decomposed module.
  • BackLite: Wraps Flash Attention 3 and uses attention sparsity to speed up the backward pass via gradient approximation.

Products

Applications and platforms built by MoonMath that leverage these acceleration technologies.

  • WorldJen: High-performance benchmarks for video and world models.
  • StoryNote — Build Worlds. Frame by Frame.

Links

Website: https://moonmath.ai

Blog: https://moonmath.ai/posts

GitHub Organization: https://github.com/moonmath-ai

X: https://x.com/moonmathai

LinkedIn: https://www.linkedin.com/company/moonmath-ai

YouTube: https://www.youtube.com/@MoonMath_ai

Popular repositories Loading

  1. LiteAttention LiteAttention Public

    Transforming Video Diffusion with Temporal Sparse Attention

    Python 46 4

  2. LiteFFN LiteFFN Public

    LiteFFN is a high-performance GPU implementation of the transformer feed-forward network (FFN) block, designed to accelerate large-scale AI workloads such as video diffusion and world models

    4

  3. surveyor2-docs surveyor2-docs Public

    Video evaluation and comparison toolkit

    Shell 3

  4. cosmos-predict2.5-moon cosmos-predict2.5-moon Public

    Python 3

  5. moondex moondex Public

    MoonMath.ai’s index of knowledge

    2

  6. Wan2.2 Wan2.2 Public

    Python 1

Repositories

Showing 10 of 22 repositories

People

This organization has no public members. You must be a member to see who’s a part of this organization.

Top languages

Loading…

Most used topics

Loading…