{"id":85866,"topic":"ai","source":"Sohu","title":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate - Sohu","url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","url_hash":"40ccc06e50a5bbfe06210a5a7bc18294b0c5e82d","author":"","summary":"<a href=\"https://news.google.com/rss/articles/CBMiiAFBVV95cUxPUHBDVXJnU3ZYMmRhdG5pNTlCR2lYZjQ2Y2w4MmVmVE9FUlJGMU5lOUdjeFRkTFV5ZVlZMHRjYkQ4R3NFZllPc19QMFdrR0c3TmI3TVhOcy1SVXNtMjZMX3F2QzVGclc4Sk1QUEhxZlBkU3Y2RDVoX2J6aWsybGRPQnFwNlcxeFNX?oc=5\" target=\"_blank\">阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">Sohu</font>","content":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","image_url":"//q9.itc.cn/q_70/images03/20260919/cf20a6ff2c45409c96512fb210787726.jpeg","lang":"zh","published_at":"2026-09-19T07:14:27+00:00","fetched_at":"2026-09-19T14:15:08+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","cluster_id":3502361,"extract_retries":0,"extract_error":null,"contract_version":"news_item.v1","format_contract_version":"news_item_formats.v1","dedup_url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 490 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":490,"summary_length":490,"usable_text_length":490,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":490,"summary_length":490}},"news_item":{"id":85866,"canonical_url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source_url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","title":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate - Sohu","source_name":"Sohu","author":null,"published_at":"2026-09-19T07:14:27+00:00","locale":"zh","topic":"ai","tags":[],"rss_summary":"<a href=\"https://news.google.com/rss/articles/CBMiiAFBVV95cUxPUHBDVXJnU3ZYMmRhdG5pNTlCR2lYZjQ2Y2w4MmVmVE9FUlJGMU5lOUdjeFRkTFV5ZVlZMHRjYkQ4R3NFZllPc19QMFdrR0c3TmI3TVhOcy1SVXNtMjZMX3F2QzVGclc4Sk1QUEhxZlBkU3Y2RDVoX2J6aWsybGRPQnFwNlcxeFNX?oc=5\" target=\"_blank\">阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">Sohu</font>","full_text":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","excerpt":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 490 characters.","diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1078215945_114760%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 490 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":490,"summary_length":490,"usable_text_length":490,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":490,"summary_length":490}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate - Sohu","url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","summary":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","source":"Sohu","date":"2026-09-19T07:14:27+00:00","content":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1078215945_114760%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 490 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 490 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":490,"summary_length":490,"usable_text_length":490,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":490,"summary_length":490}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/85866","export_markdown":"/api/items/85866/export?format=markdown","export_json":"/api/items/85866/export?format=json","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1078215945_114760%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"},"formats":{"full":{"id":85866,"title":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate - Sohu","url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source":"Sohu","author":null,"published_at":"2026-09-19T07:14:27+00:00","locale":"zh","topic":"ai","tags":[],"excerpt":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","full_text":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","reading_time_min":1,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 490 characters.","diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1078215945_114760%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 490 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":490,"summary_length":490,"usable_text_length":490,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":490,"summary_length":490}}},"quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 490 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":490,"summary_length":490,"usable_text_length":490,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":490,"summary_length":490}},"actions":{"read":"/item/85866","export_markdown":"/api/items/85866/export?format=markdown","export_json":"/api/items/85866/export?format=json","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1078215945_114760%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"}},"digest":{"id":85866,"title":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate - Sohu","url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source":"Sohu","topic":"ai","published_at":"2026-09-19T07:14:27+00:00","excerpt":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate IT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。 官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中： - 实时说话人分离，每句话归属清晰，音色复刻更稳定； - 原文译文同帧同出，双语同屏； -…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 490 characters.","reading_time_min":1,"cluster_id":3502361},"card":{"display_title":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate - Sohu","subtitle":"Sohu · 2026-09-19","summary":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate IT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。 官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate…","badges":["quality:high"],"links":{"read":"/item/85866","original":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1078215945_114760%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"},"quality_warning":null},"export":{"title":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate - Sohu","url":"https://m.sohu.com/a/1078215945_114760?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","summary":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","source":"Sohu","date":"2026-09-19T07:14:27+00:00","content":"阿里千问发布同声传译大模型 Qwen3.8-LiveTranslate\nIT之家 9 月 19 日消息，阿里千问今日发布同声传译大模型 Qwen3.8-LiveTranslate ，模型以 Interleave 架构重构实时同传，准确度、流畅度、简洁度全面提升，字均延迟（LAAL）从 2.8 秒降至 2.3 秒。\n官方表示，在支持 60 种语言的基础上，Qwen3.8-LiveTranslate 新增三项能力让同传更广泛的应用于真实场景中：\n- 实时说话人分离，每句话归属清晰，音色复刻更稳定；\n- 原文译文同帧同出，双语同屏；\n- 长上下文消歧，联系前文读懂当下，人名术语翻译更精准。\nQwen3.8-LiveTranslate 采用基于 Hybrid MoE 的 Thinker–Talker 双模块设计，通过 Interleave 方式衔接流式理解、文本输出与语音生成。其中，Thinker 把视频、音频、原文与译文编排进同一条因果序列，按时序交替排列、端到端产出，让理解与翻译在单一序列内完成；Talker 再结合译文和源音频，把译文合成为保留原说话人音色的译文语音。","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1078215945_114760%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 490 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 490 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":490,"summary_length":490,"usable_text_length":490,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":490,"summary_length":490}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}