DataLearner 标志

Gemini 3.8 FlashvsGemini 3.7 Flash

在 25 个共同 benchmark 中,Gemini 3.8 Flash 整体领先:Gemini 3.8 Flash 领先 20 项,Gemini 3.7 Flash 领先 5 项,持平 0 项,平均分差 +2.14。

Google DeepMind
Gemini 3.8 Flash

Google DeepMind · 2026-09-02 · 多模态大模型

Google DeepMind
Gemini 3.7 Flash

Google DeepMind · 2026-08-13 · 多模态大模型

Gemini 3.8 Flash20 项(80%)(20%)5 项Gemini 3.7 Flash

评测分数

按能力类目分组,每组内按分差大小排列;共 25 项。

生物与基因

Gemini 3.8 Flash 领先 3/3
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
BioMysteryBench (Human-difficult)56.501 / 2Thinking Medium (With Tools + Internet)43.502 / 2Thinking Medium (With Tools + Internet)+13
LABBench286.202 / 3Thinking Medium (With Tools + Internet)82.103 / 3Thinking Medium (With Tools + Internet)+4.10
BioMysteryBench (Human-solvable)88.801 / 2Thinking Medium (With Tools + Internet)87.102 / 2Thinking Medium (With Tools + Internet)+1.70

临床工作与医疗决策

胶着 2/2
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
MedCode48.1322 / 64Thinking High (With Tools)53.397 / 64Thinking High (With Tools)-5.26
MedScribe84.5023 / 66Thinking High (With Tools)83.9427 / 66Thinking High (With Tools)+0.56

客服与业务流程

胶着 2/2
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
SAGE35.0648 / 64Thinking High (With Tools)49.2316 / 64Thinking High (With Tools)-14.17
τ³-Banking45.8014 / 167Thinking Medium (With Tools)35.5046 / 167Thinking Medium (With Tools)+10.30

法律

Gemini 3.8 Flash 领先 2/2
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Legal Research Bench38.9421 / 42Thinking High (With Tools)34.6228 / 42Thinking High (With Tools)+4.32
Harvey's Legal Agent Benchmark1011 / 43Thinking High (With Tools)8.7513 / 43Thinking High (With Tools)+1.25

跨能力聚合指数

Gemini 3.8 Flash 领先 2/2
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Vals Index v2.154.8314 / 36Thinking High (With Tools)51.2723 / 36Thinking High (With Tools)+3.56
Vals Index62.2510 / 42Thinking High (With Tools)59.3116 / 42Thinking High (With Tools)+2.94

金融

Gemini 3.8 Flash 领先 2/2
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Tax Agent Bench66.779 / 20Thinking High (With Tools)57.6619 / 20Thinking High (With Tools)+9.11
EMB (Excel Modeling)72.206 / 42Thinking High (With Tools)71.339 / 42Thinking High (With Tools)+0.87

写作与创意表达

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Creative Writing1,74826 / 110Normal (No Tools)1,72328 / 110Normal (No Tools)+24.70

图像感知与视觉推理

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
MMMU-Pro85.604 / 229Thinking High (No Tools)85.505 / 229Thinking High (No Tools)+0.10

抽象归纳与泛化

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
ARC-AGI-198.501 / 193Thinking High (No Tools)95.5032 / 193Thinking High (No Tools)+3

数学

Gemini 3.7 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
ProofBench v1.1 (Lean 4)4820 / 28Thinking High (With Tools)5813 / 28Thinking High (With Tools)-10

文档与图表理解

Gemini 3.7 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
GDP.pdf22.8026 / 122Thinking Medium (No Tools)341 / 122Thinking Medium (No Tools)-11.20

科学知识与推理

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
CritPt18.3045 / 204Thinking High (No Tools)14.3061 / 204Thinking High (No Tools)+4

科学计算与科研编程

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Terminal-Bench-Science 0.112.4010 / 18High Thinking (With Tools)5.7016 / 18High Thinking (With Tools)+6.70

程序生成与编辑

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Vibe Code Bench v1.178.6519 / 61Thinking High (With Tools)70.3926 / 61Thinking High (With Tools)+8.26

算法与竞赛编程

Gemini 3.7 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
IOI (Vals v2)56.9412 / 26Thinking High (With Tools)67.839 / 26Thinking High (With Tools)-10.89

维护、优化与工程管理

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Code Migration36.5524 / 43Thinking High (With Tools)34.8026 / 43Thinking High (With Tools)+1.75

综合知识考试

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
HLE-Verified54.901 / 6Thinking Medium (No Tools)53.604 / 6Thinking Medium (No Tools)+1.30

自主开发与终端任务

Gemini 3.8 Flash 领先 1/1
评测项Gemini 3.8 FlashGemini 3.7 Flash分差
Terminal-Bench 2.189.404 / 57Thinking (With Tools)85.8017 / 57Thinking (With Tools)+3.60

规格对比

字段Gemini 3.8 FlashGemini 3.7 Flash
发布机构Google DeepMindGoogle DeepMind
发布时间2026-09-022026-08-13
模型类型多模态大模型多模态大模型
架构稠密模型稠密模型
参数规模暂无数据暂无数据
上下文长度1M1M
最大输出64K64K

API 调用价格

价格优先使用 DataLearner 配置的 API 记录;缺失项不做推测。

价格项Gemini 3.8 FlashGemini 3.7 Flash
文本输入$0.75 / 1M tokens$0.75 / 1M tokens
文本输出$3.75 / 1M tokens$3.75 / 1M tokens
缓存读取$0.075 / 1M tokens$0.075 / 1M tokens

小结

  • Gemini 3.8 Flash在以下类目领先:生物与基因 (3/3)、法律 (2/2)、跨能力聚合指数 (2/2)、金融 (2/2)、写作与创意表达 (1/1)、图像感知与视觉推理 (1/1)、抽象归纳与泛化 (1/1)、科学知识与推理 (1/1)、科学计算与科研编程 (1/1)、程序生成与编辑 (1/1)、维护、优化与工程管理 (1/1)、综合知识考试 (1/1)、自主开发与终端任务 (1/1)
  • Gemini 3.7 Flash在以下类目领先:数学 (1/1)、文档与图表理解 (1/1)、算法与竞赛编程 (1/1)
  • 胶着类目:临床工作与医疗决策、客服与业务流程

25 个共同 benchmark 上,Gemini 3.8 Flash 平均高出 2.14 分。

单项差距最大的 benchmark:Creative Writing — Gemini 3.8 Flash 1,748,Gemini 3.7 Flash 1,723(分差 +24.70)。

本页正文由结构化模型、价格与 benchmark 数据生成,不使用实时 LLM 撰写。