LLaVA-OneVision-1.5

Unified multimodal foundation spanning images, multi-image, and video understanding in one open model.

Get Started Learn More

LMMs-Eval v0.4

Comprehensive evaluation framework for multimodal models.

LLaVA-Critic-R1

Unified critic and policy model through reinforcement learning at the 7B scale.

Aero-1-Audio

Frontier Open Audio Models.

About Us

LMMs-Lab is a non-profit research-oriented organization with a group of passionate researchers, we share the sincere passion for developing multimodal intelligence.

Research Areas

Models

Access a growing collection of state-of-the-art open-source multimodal models.

Browse

Tools

Utilize intuitive tools designed for efficient development and experimentation.

Explore

Research

Explore cutting-edge research and contribute to the advancement of multimodal AI.

Dive in