Naver和SOOP为直播平台加入实时AI字幕、翻译和切片功能

Real-time streaming platform companies are accelerating service differentiation by integrating artificial intelligence (AI) technology across content production and viewing experiences. As language barriers fall and content accessibility improves, the global user base of domestic streaming platforms is also expected to expand.
According to the information technology (IT) industry on the 4th, Naver's "Chzzk" and "SOOP" have recently introduced new AI-based features to coincide with the live broadcast schedules of major international tournaments.
Chzzk fully upgraded its AI caption service on the occasion of its exclusive broadcast of the "Esports World Cup (EWC 2026)," held in Paris, France, last month. Previously, it used a sequential "cascade" method that converted speech into text and then had AI translate and process it. However, this process had the limitation of losing much of the non-verbal information, such as the speaker's emotions and intonation. To overcome this limitation, Naver introduced a large language model (LLM)-based "end-to-end (E2E)" model. The new model grasps the entire context of the speech to reflect even subtle nuances, and processes highly difficult technical terms more accurately.
Earlier, during the live broadcast of the 2026 FIFA World Cup, Naver introduced "AI Clip" and "AI Chapter" features to Chzzk. AI Clip is a service that detects decisive moments, such as goal scenes during a match, in real time and automatically generates them into short-form videos. AI Chapter automatically analyzes and divides video content by segment, helping users easily select and watch the match scenes they want.
SOOP is also advancing its AI features to strengthen communication among global users. On the 4th, SOOP improved its chat translation function so that conversations can continue in one's native language without any separate button operation. As a result, communication is expected to become more active—centered on broadcasts of global esports tournaments such as "League of Legends (LoL)"—not only between streamers and users of different nationalities but also among users.
SOOP also introduced a caption service combining real-time speech-to-text (STT) and AI translation technology, so that overseas users who are not provided with broadcasts in their local language can easily enjoy the content. It has built an environment in which overseas users without broadcasts in their local language can readily enjoy the content. AI captions were also applied to the Seoul series of the global professional dance league "International Dance League (IDL)," held on the 1st, drawing a strong response from overseas users. In addition, SOOP is operating a "broadcast summary" feature that automatically organizes key content based on broadcast audio and chat data.
While domestic streaming services have so far remained in the domestic market, the influx of global users is expected to grow as AI technology is actively applied. "AI technologies such as real-time translation, captions and automatic highlight generation not only improve content accessibility but also play a key role in expanding the global user base," an industry official said. "Going forward, creating an environment where users from various countries can enjoy a single piece of content simultaneously will become a platform's core competitiveness."