Step3: Cost-Effective Multimodal Intelligence

Step3 is our cutting-edge multimodal reasoning model—built on a massive Mixture-of-Experts architecture with 321 billion total parameters and 38 billion active. It is designed end-to-end to minimize decoding costs while delivering top-tier performance in vision–language reasoning, mathematics, and code. Through the co-design of Multi-Matrix Factorization Attention (MFA) and Attention-FFN Disaggregation (AFD), Step3 maintains exceptional efficiency across both flagship and low-end accelerators.

Links: GitHub | Project Page | Paper

OpenCoder: The Open Cookbook for Top-Tier Code Large Language Models

OpenCoder is an open and reproducible code LLM family which includes 1.5B and 8B base and chat models, supporting both English and Chinese languages. Starting from scratch, OpenCoder is trained on 2.5 trillion tokens composed of 90% raw code and 10% code-related web data, reaching the performance of top-tier code LLMs. We provide not only model weights and inference code, but also the reproducible training data, the complete data processing pipeline, rigorous experimental ablation results, and detailed training protocols.

Links: GitHub | HuggingFace | Project Page | Paper

INF-34B: INF's Open-Source Large Language Models

INF-34B has 34 billion parameters with a context window length of 32K, and is trained on about 3.5T well-processed tokens from English and Chinese bilingual corpus. Compared with open source models of comparable size, INF-34B not only provides competitive performance in the OpenCompass evaluation, but also has impressive potential in both finance and healthcare domains. Besides, the quantized INF-34B runs on graphics cards of 24GB VRAM with negligible accuracy loss, which facilitates commercial applications, especially low-resource scenarios.

Links: GitHub | HuggingFace | Tech Report