Thinking Machines 推出原生互動模型 TML-Interaction-Small 276B-A12B：突破即時語音 SOTA 並淘汰傳統 VAD
Original: [AINews] Thinking Machines' Native Interaction Models - TML-Interaction-Small 276B-A12B - advances SOTA Realtime Voice and kills standard VAD
According to AINews, the AI research team Thinking Machines (affectionately nicknamed "Team Thinky" by the community) has recently unveiled…
AI 研發團隊 Thinking Machines 推出全新原生互動模型「TML-Interaction-Small 276B-A12B」。該模型專為即時語音設計，擁有 276B 總參數與 12B 啟用參數。它不僅刷新了即時語音的技術前沿（SOTA），更透過原生互動機制，徹底取代了傳統的語音活動檢測（VAD）模組，實現更自然、無縫的雙向語音對話。
According to AINews, the AI research team Thinking Machines (affectionately nicknamed "Team Thinky" by the community) has recently unveiled a new native interaction model — **TML-Interaction-Small 276B-A12B**. This model achieves a breakthrough in the field of real-time voice interaction and is considered likely to render traditional Voice Activity Detection (VAD) technology obsolete.
Free shows the 3-line summary; Pro unlocks the full deep summary (~300 words) so you never have to click through.
See Pro plans →Want the original English / full article?
Read on Latent Space →Summaries are AI-generated; the original article is authoritative.