IndexTTS2 is a next-generation text-to-speech model developed by Bilibili, officially open-sourced on September 8, 2025. The model achieves major breakthroughs in emotional expression and duration control, being hailed by the community as 'the most realistic and expressive TTS model.'
Qwen3-ASR-Flash is a next-generation speech recognition service developed by Alibaba's Tongyi Qianwen team based on the Qwen3-Omni multimodal foundation model.
Kimi K2-0905 is Moonshot AI's latest open-source large language model featuring 1 trillion parameters with 32B active, 256K context length, and exceptional coding capabilities approaching Claude Sonnet 4 level.
Tencent Hunyuan Translation Model is a professional translation AI model open-sourced by Tencent on September 1, 2025, consisting of two core components: Hunyuan-MT-7B and Hunyuan-MT-Chimera-7B.