Google

Gemini 2.5 Flash Lite Preview 09-2025 API 基准测试价格和服务商数据

Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better...

分类性能

按基准分类汇总的 0-100 分表现。

覆盖: 2 / 8
平均分: 48.0

#1综合推理

49.8

#2代码

46.3

暂无数据:智能体知识数学多语言多模态指令遵循

技术规格

Input and output token limits for this model, plus how it ranks on long-context understanding.

INPUT

1.0Mtokens

≈ 1.3K pages of text

OUTPUT

65.5Ktokens

8K128K1M4M

1.0M

能力

Technical Details

输入

排名

擅长

详细分数

更新时间: Jul 9, 2026

价格

2 个指标

输入价格$0.100/M#7 / 138输出价格$0.400/M#12 / 138

输入价格

Input price

来源: Artificial Analysis / Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)站内排名: #7 / 138更新时间: Jul 9, 2026置信度: 4

5$0.070/M

6$0.075/M

7$0.100/M

7Gemini 2.5 Flash Lite Preview 09-2025$0.100/M

7

输出价格

Output price

来源: Artificial Analysis / Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)站内排名: #12 / 138更新时间: Jul 9, 2026置信度: 4

8$0.300/M

12$0.400/M

推理与知识

3 个指标

MMLU-Pro79.6%#60 / 113GPQA65.1%#103 / 152

MMLU-Pro

来源: Artificial Analysis / Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)站内排名: #60 / 113更新时间: Jul 9, 2026置信度: 4

5680.2%

代码与智能体

2 个指标

LiveCodeBench64.1%#41 / 108SciCode28.5%#115 / 149

LiveCodeBench

来源: Artificial Analysis / Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)站内排名: #41 / 108更新时间: Jul 9, 2026置信度: 4

3767.2%

相关模型对比

GGemini 2.5 Flash Lite Preview 09-2025 vs GPT-5.4

3 个可验证比较点

可索引

GGemini 2.5 Flash Lite Preview 09-2025 vs GPT-5

3 个可验证比较点

可索引

CClaude Opus 4.6 vs Gemini 2.5 Flash Lite Preview 09-2025

3 个可验证比较点

可索引

数据更新于 2026年7月9日 09:37排名基于社区提交的测试数据与定期健康探测，仅供参考，非官方数据。Artificial Analysis标准基准数据可能包含 BenchLM 等公开来源。

输出

知识截止

2025-01-31

官方文档

Tokenizer

Gemini

架构

text+image+file+audio+video->text

内容审核

否

到期日期

2026-07-09

支持参数

include_reasoningmax_tokensreasoningresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_p

GPT-4.1 Nano

$0.100/M

7MiMo-V2-Flash$0.100/M

7Qwen3.5 Omni Flash$0.100/M

7Qwen3.5-9B$0.100/M

7Step 3.5 Flash$0.100/M

12Gemini 2.5 Flash Lite Preview 09-2025$0.400/M

12GLM-4.7 Flash$0.400/M

12GPT-4.1 Nano$0.400/M

12GPT-5 Nano$0.400/M

17DeepSeek V3.2$0.420/M

17DeepSeek V3.2 Exp$0.420/M

57

Claude Haiku 4.5

80.0%

58Grok 379.9%

59Qwen3 32B79.8%

60Gemini 2.5 Flash Lite Preview 09-202579.6%

61R1 Distill Llama 70B79.5%

62GLM-4.779.4%

63Ring Flash 2.079.3%

64Claude Sonnet 4.679.2%

65O3 Mini79.1%

GPQA

来源: Artificial Analysis / Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)站内排名: #103 / 152更新时间: Jul 9, 2026置信度: 4

99GPT-4.166.6%

100GLM-4.766.4%

100GPT-4.1 Mini66.4%

102Qwen365.9%

103Gemini 2.5 Flash Lite Preview 09-202565.1%

10464.6%

10563.6%

10662.5%

10762.2%

10861.6%

HLE

来源: Artificial Analysis / Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)站内排名: #110 / 150更新时间: Jul 9, 2026置信度: 4

105Grok 35.1%

107Gemini 1.5 Pro4.9%

107O1 Mini4.9%

109Gemini 2.0 Flash4.7%

110Gemini 2.5 Flash Lite Preview 09-20254.6%

1104.6%

1144.4%

38GPT-5.266.9%

39gpt-oss-20b65.2%

40Qwen3 Max65.1%

41Gemini 2.5 Flash Lite Preview 09-202564.1%

42Ring Flash 2.062.8%

43Qwen3 235B A22B62.2%

44Kimi K2 090561.0%

45Qwen3 VL 235B A22B Instruct59.4%

46Gemini 2.5 Flash Lite59.3%

SciCode

来源: Artificial Analysis / Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)站内排名: #115 / 149更新时间: Jul 9, 2026置信度: 4

111Qwen3 VL 32B Instruct30.1%

112Gemini 1.5 Pro29.5%

113Qwen3.5-35B-A3B29.3%

114Olmo 3 32B Think28.6%

115Gemini 2.5 Flash Lite Preview 09-202528.5%

11528.5%

11827.8%

12027.7%

GLM-4.7 Flash

Nemotron 3 Nano Omni

Devstral Small

Gemini 2.5 Flash Lite

MiMo-V2-Flash

Nemotron 3 Nano Omni

Step 3.5 Flash

Gemini 2.5 Flash Lite

Gemini 2.5 Flash Lite

Gemini 2.0 Flash Lite

Qwen3 Coder 30B A3B Instruct

Qwen3.5-9B