A 0.1B Omni model trained from scratch
MOSS-TTS-Nano is an open-source multilingual tiny speech generation
Workplace AI platform for enterprise search and workflow automation
Any model. Any hardware. Zero compromise
OCR model for complex documents with layout-aware structured outputs
Shared repository for open-sourced projects from the Google AI Lang
Numerical differential equation solvers in JAX
Unsupervised Learning for Image Registration
Python Audio Analysis Library: Feature Extraction, Classification
Build multimodal AI applications with cloud-native stack
Minimal reproduction of OneRec
Visual intelligence for your home.
From nobody to big model (LLM) hero
NeurIPS2025 Spotlight] Quantized Attention
Open-source evaluation toolkit of large multi-modality models (LMMs)
RAG Search API
Qwen3-ASR is an open-source series of ASR models
Spark-TTS Inference Code
Foundation model for image generation
Marrying Grounding DINO with Segment Anything & Stable Diffusion
OpenTinker is an RL-as-a-Service infrastructure for foundation models
Motion-controllable Video Generation via Latent Trajectory Guidance
A tool to use the Ai2 Open Coding Agents Soft-Verified Agents
Multimodal embedding and reranking models built on Qwen3-VL
A New Axis of Sparsity for Large Language Models