Seedream 4.5
Search documents
豆包 1.8 多模态超越谷歌Gemini 3!字节祭出“推理代工”,要做模型届的英特尔?
AI前线· 2025-12-18 07:24
运行效率方面,豆包 1.8 的 TPM(每分钟处理 Token 数)达到了 5000k,而 RPM(每分钟请求数)为 30k。 谭待表示,这些能力让豆包 1.8 在 AIME 2025 等 Agent 评测集上稳步提升,在通用智能体测评 Benchmark 上取得全球领先成绩;在多模态理解 方面,于视觉判断准确性、空间理解、文档解析、视频运动识别等多项能力上超越 Gemini 3,在其他任务上也都处于全球第一梯队水平。 | | | | | 豆包大模型1.8 | | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | | | | | | | | | Qwen3-235b-a22b- | | | 紹力推度 | 测浮集 | GPT-5 High | Claude-Sonnet-4.5 | Gemini-3-pro | Doubao Seed-1.8 | Doubao Seed-1.6-1015-high | thinking | | | 数学 | AIME-25 | 94.6 | 87.0 | 95.0 | 94.3 | 85.0 | 87 ...
Nano Banana平替悄悄火了!马斯克、Meta争相合作
Sou Hu Cai Jing· 2025-12-15 10:57
| 7 | Creator 14 | Model 17 | ELO 1. | 95% CI | Appearances 1J | Release Date 14 | | --- | --- | --- | --- | --- | --- | --- | | 1 G Google | | Nano Banana Pro (Gemini 3 Pro Image) | 1,223 | -10/+10 | 5,921 | Nov 2025 | | 2 | & Black Forest Labs | FLUX.2 [pro] | 1,195 | -10/+11 | 4,852 | Nov 2025 | | ന | In ByteDance Seed | Seedream 4.0 | 1,193 | -6/+6 | 16,805 | Sept 2025 | | | & Black Forest Labs | FLUX.2 [flex] 0 | 1,182 | -11/+11 | 4,841 | Nov 2025 | | ន | Iil ByteDance Seed | Seedream 4.5 | 1,170 | -9 ...
国信证券晨会纪要-20251209
Guoxin Securities· 2025-12-09 01:01
证券研究报告 | 2025年12月09日 | 晨会纪要 | | --- | | 数据日期:2025-12-08 | 上证综指 | 深证成指沪深 | 300 指数 | 中小板综指 | 创业板综指 | 科创 50 | | --- | --- | --- | --- | --- | --- | --- | | 收盘指数(点) | 3924.07 | 13329.99 | 4621.75 | 14193.95 | 3889.74 | 1350.79 | | 涨跌幅度(%) | 0.54 | 1.38 | 0.81 | 0.93 | 2.12 | 1.86 | | 成交金额(亿元) | 8394.39 | 11972.07 | 4967.16 | 3786.42 | 5818.21 | 639.51 | $\frac{\pi\times6}{\pi\times6}$. (1998) (200) 宏观与策略 宏观专题:宏观经济专题研究-政治钟摆下的美联储:FOMC 票委换届前瞻 宏观快评:12 月政治局会议解读-提质增效,内需主导 固 定 收 益 快 评 : 可 交 换 私 募 债 跟 踪 - 私 募 EB 每 周 跟 ...
DeepSeek-V3.2和豆包手机助手解读
Guotou Securities· 2025-12-07 12:08
Investment Rating - The report maintains an investment rating of "Outperform the Market - A" [7] Core Insights - DeepSeek has launched the V3.2 model, enhancing its reasoning capabilities to a globally leading level, suitable for everyday use in Q&A and general agent tasks [12][27] - The V3.2 model achieved performance comparable to GPT-5 in benchmark tests, slightly below Gemini-3.0-Pro, while significantly reducing output length and computational costs [12][27] - The introduction of the DSA (DeepSeek Sparse Attention) mechanism reduces context computation costs, changing complexity from O(L²) to O(Lk), where k is a fixed value of 2048 [13][14] - The report highlights the launch of the Doubao mobile assistant, which integrates AI capabilities into mobile operating systems, allowing users to perform complex tasks with voice commands [15] Summary by Sections Industry Performance - The computer sector underperformed relative to the Shanghai Composite Index, with a 1-month relative return of -5.4% and a 3-month return of -4.5% [5][16] - The computer sector index ranked 25th among 30 industry indices, indicating weaker performance [19] Important Industry News - Google’s TPUv7 has begun to challenge NVIDIA's dominance in AI chips, marking a significant shift in the competitive landscape [25] - The 2025 World Computing Conference showcased advancements in computing systems, emphasizing the importance of system capabilities over individual card performance [26]