Google DeepMind 新负责人称 Gemini 4 即将准备就绪
英文标题：Gemini 4 is almost ready, says new Google DeepMind chief
据报道，Google 正接近推出备受期待的 Gemini 4，此前其旗舰 AI 模型发布节奏已落后于竞争对手。Google DeepMind 新负责人 Koray Kavukcuoglu 表示，Gemini 4 目前处于完善阶段，公司计划在远早于年底的时间推出早期后训练版本。自 2025 年 11 月发布 Gemini 3 系列以来，Google 尚未推出新的旗舰模型，而 OpenAI GPT-6 和 Anthropic Mythos 已先后上线并在性能上领先 Gemini 3。Google 此前选择优先推出更快、能力较弱的 Flash 模型，而非发布 Gemini 3.5 Pro；Kavukcuoglu 则坚称，Google 将始终处于 AI 技术前沿。
34.6 KB
86.1 KB
3.1 KB
207.5 KB
315.2 KB
307.7 KB
490.0 KB
1.2 MB
100.1 KB
158.8 KB
416.5 KB
249.8 KB
807.7 KB
1.5 MB
95.0 KB
40.4 KB
中文
据报道，Google 正接近推出其备受期待的 Gemini 4 模型，此前该公司在旗舰 AI 模型发布方面落后于竞争对手。Google DeepMind 部门负责人 Koray Kavukcuoglu 在就任后的首次媒体亮相中接受了 The Information 采访。他表示，Gemini 4 目前正处于完善阶段，公司目标是在远早于年底的时间推出该模型。
Kavukcuoglu 对该媒体表示：“我们的意图是，嗯——尽可能快地——发布一个早期的后训练输出版本，因为我们看到了结果，而且对此感到兴奋。”他还补充称，Google 此后将“持续进行快节奏迭代”。在 DeepMind 高层领导团队调整后，这有望标志着 Google 重回更稳定的 AI 发布节奏。前 AI 负责人 Demis Hassabis 已于 8 月卸任，此前他曾表示，Gemini 4 已取得“重大进展”。
自 2025 年 11 月发布 Gemini 3 系列 以来，Google 尚未发布新的旗舰 AI 模型。此后，OpenAI 的 GPT-6 和 Anthropic 的 Mythos 系列模型均已推出。两者的性能都优于 Gemini 3，使这两家竞争性 AI 提供商在 AI 竞赛中对 Google 建立了显著领先优势。Google CEO Sundar Pichai 曾在 5 月的 I/O 活动上表示，重大 Gemini 3.5 Pro 更新将于 6 月推出，以帮助缩小差距，但该模型始终未能上线。
Kavukcuoglu 表示，公司“稍微退了一步”，将重点放在推出速度更快、能力较弱的 Flash 模型，而不是发布 Gemini 3.5 Pro。当被问及是否担心 Google 在 AI 研发上落后于竞争对手时，他告诉 The Information：“在我看来，我们始终会处于前沿，这一点是确定无疑的。”
English / 原文
Google is reportedly nearing the launch of its long awaited Gemini 4 model, after dawdling behind rival developers on flagship AI releases. Speaking with The Information during his first media appearance as the leader of Google’s DeepMind division, Koray Kavukcuoglu said that Gemini 4 is currently in its refinement stage, and that the company is aiming to launch the model “much earlier” than the end of the year.
“Our intention is to, like — as soon as possible — to release an early post-training output because we see the results and we are excited,” Kavukcuoglu told the publication, adding that Google would “continue the fast-paced iterations” going forward. This hopefully marks Google returning to a more consistent AI release timeline following a shakeup at DeepMind’s top leadership, with former AI head Demis Hassabis stepping down in August after announcing that “great progress” was being made on Gemini 4.
Google hasn’t released a new flagship AI model since the Gemini 3 series in November 2025. OpenAI’s GPT-6 and Anthropic’s Mythos series models have launched since then. Both outperform Gemini 3 and give the rival AI providers a sizable lead against Google in the AI race. Google CEO Sundar Pichai said at the I/O event in May that a major Gemini 3.5 Pro update would be coming in June to help bridge the gap, but the model was never rolled out.
Kavukcuoglu said the company “took a little bit of a step back” to focus on introducing faster, less powerful Flash models instead of releasing Gemini 3.5 Pro. When asked whether he was concerned about Google falling behind its rivals on AI development, he told The Information that “In my mind, it’s a certainty that we are always gonna be at the frontier.”