Youdao open-sources Confucius 4 multimodal LLM, cuts costs

NetEase Youdao has announced a significant upgrade to its "Confucius 4" large language model, now entering the multimodal era with support for text, image, and audio interactions. The company is open-sourcing its core multimodal and text-to-speech (TTS) models, aiming to reduce implementation costs for developers. The new model demonstrates state-of-the-art performance in visual mathematical reasoning and offers a 43.2% reduction in reasoning chain output length, leading to lower inference costs. AI

影响 Lowers barriers for developers in multimodal and speech synthesis, potentially accelerating AI Agent development and adoption.

排序理由 This is a significant product release and open-source initiative from a major tech company in the AI space. [lever_c_demoted from significant: ic=1 ai=1.0]

在 雷峰网 (Leiphone) 阅读 →

• Confucius 4
• NetEase Youdao
• text-to-speech (TTS) model
• multimodal model
• translation model

• 模型发布
• 产品
• 其他

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →