跳转到主要内容

Gemini 2.5 Flash Preview (Reasoning)

GoogleGeminiProprietary

描述

A thinking model designed for a balance between price and performance. It builds upon Gemini 2.0 Flash with upgraded reasoning, hybrid thinking control, multimodal capabilities (text, image, video, audio input), and a 1M token input context window.

发布日期
2025-04-17
参数规模
上下文长度
1.0M
支持模态
audio, image, pdf, text, video

能力雷达图

36
general
47
coding
86
reasoning
46
science
74
agents
80
multimodal

排行榜排名

领域#排名分数来源
代码能力榜193
55.0
AA
通用能力榜269
47.0
AA
科学能力244
50.0
AA

基准测试分数 (LLM Stats)

(LLM Stats (zeroeval))

Biology

GPQANYU + Cohere + Anthropic (2023)82.8%自报

Code

Aider-Polyglot61.9%自报
SWE-Bench Verified60.4%自报
Aider-Polyglot Edit56.7%自报

Factuality

FACTS Grounding85.3%自报
SimpleQA26.9%自报

General

Global-MMLU-Lite88.4%自报
MMMU79.7%自报
Vibe-Eval65.4%自报
LiveCodeBench v563.9%自报
MRCR32.0%自报

Math

AIME 202488.0%自报
AIME 202572.0%自报
Humanity's Last Exam11.0%自报

AA 评测指数

(Artificial Analysis)
Intelligence Index(Artificial Analysis)
17.7
Math 500(OpenAI (2024), subset of Hendrycks et al. MATH (2021))
1.0
Aime(MAA (Mathematical Association of America))
0.8
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))
0.8
Gpqa(NYU + Cohere + Anthropic (2023))
0.7
Livecodebench(UC Berkeley + MIT + Cornell (2024))
0.5
Scicode(UIUC + Argonne National Lab (2024))
0.4
Hle(Center for AI Safety + Scale AI (2025))
0.1

LLM Stats 分类评分

(LLM Stats (zeroeval))
Language
90
Grounding
90
Physics
80
Healthcare
80
Biology
80
Chemistry
80
Multimodal
70
Math
60
Reasoning
60
Factuality
60
Frontend Development
60
General
60
Code
60
Vision
50
Long Context
20

定价

输入价格免费
输出价格免费
混合价格(3:1)免费
缓存读取价格$0.03 / 1M tokens

速度

Tokens/秒0.0
首Token延迟0.00s
首回答延迟0.00s

供应商价格排行

暂无提供商数据

外部链接