English(EN) DeepSeek says new Flash AI model beats Kimi K3 on cyber, coding benchmarks

DeepSeek 的 V4.1 Flash AI 模型在基准测试中超越 Kimi K3 和 GPT-5.6 Sol

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-10 09:26

DeepSeek 推出了其 V4.1 Flash 模型，这是一个新的人工智能系统，据报道在特定基准测试中优于 Moonshot AI 的 Kimi K3 和 OpenAI 的 GPT-5.6 Sol 等竞争对手。该模型采用了新颖的“因果编码器-解码器”架构和专家混合（MoE）设计，使其能够激活其 5520 亿参数中的一小部分进行处理。这种效率对于面临硬件成本和芯片出口限制的中国人工智能开发者至关重要，使他们能够以较低的运营成本提供高性能模型。 AI

影响为多模态模型设定了新的效率基准，可能影响成本受限市场的开发策略。

排序理由来自人工智能开发商的新模型发布，声称具有基准性能。[lever_c_demoted from frontier_release: ic=1 ai=1.0]

在 SCMP — Tech 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

DeepSeek 的 V4.1 Flash AI 模型在基准测试中超越 Kimi K3 和 GPT-5.6 Sol

本文如何被排名

Signal score: 41 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Significant; 来自人工智能开发商的新模型发布，声称具有基准性能。[lever_c_demoted from frontier_release: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: model release, infra; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

SCMP — Tech TIER_1 English(EN) · Chong Ming Lee · 2026-09-10 09:26

DeepSeek称其新Flash AI模型在网络和编码基准测试中超越Kimi K3

DeepSeek has released its V4.1 Flash model, claiming it outperforms its previous flagship while cutting inference costs and boosting speeds – the latest salvo in China’s aggressive price-and-performance war. The Chinese artificial intelligence developer said on Thursday that V4.1…

本文如何被排名

报道来源 [1]

DeepSeek称其新Flash AI模型在网络和编码基准测试中超越Kimi K3

相关实体

相关话题