YuanXiang Open Source High-Performance Large Model XVERSE-65B-Chat Edition Unconditionally Free for Commercial Use


Academic team releases OpenSeeker-v2, breaking industry dominance in deep search. It achieves top-tier agent capabilities with high-quality data, bypassing resource-heavy pipelines (pretraining, CPT, SFT, RL), offering a new paradigm for LLMs.....
At the 9th Digital China Construction Summit, China Mobile will publicly showcase its self-developed "Jiutian" 35B general large model for the first time. Moortech announced that its flagship GPU MTTS 5000 has completed full-process adaptation and inference verification for this model, marking an important breakthrough in the domestic computing power ecosystem. Through its self-developed MUSA software stack and the high-performance inference engine SGLang-MUSA, Moortech has achieved deep integration and optimization throughout the entire inference pipeline of the "Jiutian" 35B model.
DeepSeek announced a significant reduction in the price of its entire series of API input cache to 1/10 of the initial price. The input cache price for V4-Pro is now reduced to 0.1 yuan per million Tokens, and after a limited-time promotion, it is only 0.025 yuan, far lower than the overseas competitor GPT-5.5Pro. This move aims to attract more developers and enterprises by offering extreme cost-effectiveness, marking a new stage in domestic AI cost control.
Xiaomi released the MiMo-V2.5 series of large models on April 23 and initiated public testing. The series includes four models, with the core models MiMo-V2.5-Pro and MiMo-V2.5 being open-sourced globally, demonstrating its commitment to promoting an open AI ecosystem. This update is not only a product iteration but also a comprehensive upgrade of the technology foundation, featuring flagship performance that supports a context length of up to one million and complex task processing.
Xiaomi released the MiMo-V2.5 series of large models, including MiMo-V2.5, V2.5-Pro, and accompanying TTS and ASR models, marking an upgrade from "usable" to "user-friendly." The flagship model MiMo-V2.5-Pro has reached competitive levels with top models such as Claude Opus4.6 and GPT-5.4 in terms of general intelligent agent capabilities and software engineering. Its core advantage lies in high instruction adherence and self-correction capabilities.