OpenRouter 使用分布：语种、上下文与工具调用

自然语言分布

英语 · 2026-09-28 → 2026-10-27

按提示词语种分类，同一模型在不同语种下的用量差别明显

英语简体中文繁体中文西班牙语日语韩语俄语法语德语葡萄牙语阿拉伯语印地语

排名	模型	厂商	token 用量（周）	占比	环比
1	DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash	DeepSeek	1.74T	15.8%	+0.5%
2	Jev 1.13typesafe/jev-1.13	TypeSafe	910.0B	8.3%	新上榜
3	DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash	DeepSeek	799.9B	7.3%	-3.0%
4	GLM 5.3 Flashz-ai/glm-5.3-flash	Z.ai	793.9B	7.2%	-35.2%
5	Hy4 previewtencent/hy4-preview	Tencent	499.6B	4.6%	+2.5%
6	Step 5 Previewstepfun/step-5-preview	Stepfun	495.0B	4.5%	-33.2%
7	GPT-6 Lunaopenai/gpt-6-luna	OpenAI	443.0B	4.0%	-27.0%
8	Gpt 6 Luna Decisionsopenai/gpt-6-luna-decisions	OpenAI	336.9B	3.1%	新上榜
9	MiniMax M3.1 Flashminimax/minimax-m3.1-flash	MiniMax	323.3B	2.9%	新上榜

英语合计 10.97T，未进前 9 名的模型合计占 42.2%。

编程语言分布

Python · 2026-09-28 → 2026-10-27

按代码请求的编程语言分类，反映各语言模型的偏好度

Python JavaScript TypeScript Java C++ C# Go Rust SQL Ruby

排名	模型	厂商	token 用量（周）	占比	环比
1	DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash	DeepSeek	380.6B	16.1%	+7.5%
2	GLM 5.3 Flashz-ai/glm-5.3-flash	Z.ai	212.9B	9.0%	-25.1%
3	DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash	DeepSeek	181.4B	7.7%	+6.9%
4	Jev 1.13typesafe/jev-1.13	TypeSafe	158.0B	6.7%	新上榜
5	GPT-6 Lunaopenai/gpt-6-luna	OpenAI	110.0B	4.7%	-14.1%
6	Hy4 previewtencent/hy4-preview	Tencent	107.7B	4.6%	+11.7%
7	Gpt 6 Luna Decisionsopenai/gpt-6-luna-decisions	OpenAI	80.6B	3.4%	-9.3%
8	DeepSeek V4 Flash 0423deepseek/deepseek-v4-flash	DeepSeek	77.7B	3.3%	+10.2%
9	Step 5 Previewstepfun/step-5-preview	Stepfun	66.7B	2.8%	新上榜

Python合计 2.36T，未进前 9 名的模型合计占 41.7%。

上下文长度分布

1K – 10K · 2026-05-04 → 2026-10-26

按提示词 + 完成内容的长度分桶统计请求量

1K 以内 1K – 10K 10K – 100K 100K – 1M 1M – 10M

排名	模型	厂商	请求数（周）	占比	环比
1	Jev 1.13typesafe/jev-1.13	TypeSafe	186.2M	14.1%	-69.1%
2	GLM 5.3 Flashz-ai/glm-5.3-flash	Z.ai	115.7M	8.8%	-71.5%
3	Jev 1.14typesafe/jev-1.14	TypeSafe	89.8M	6.8%	-65.3%
4	GPT-6 Lunaopenai/gpt-6-luna	OpenAI	82.1M	6.2%	-69.3%
5	DeepSeek V4 Flash 0423deepseek/deepseek-v4-flash	DeepSeek	77.0M	5.8%	-65.8%
6	DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash	DeepSeek	72.3M	5.5%	-66.4%
7	DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash	DeepSeek	65.3M	5.0%	-67.5%
8	DeepSeek V4 Flash 0423deepseek/deepseek-v4-flash	DeepSeek	58.2M	4.4%	-69.6%
9	Claude Haiku 5.5anthropic/claude-haiku-5.5	Anthropic	29.2M	2.2%	新上榜

1K – 10K合计 1.3B，未进前 9 名的模型合计占 41.1%。

请求主题领域

近 30 天 · 17 个领域

按请求主题分类的花费份额

排名	领域	花费份额	该领域花费最高的模型（份额 / 环比）
1	软件与技术Software & technology	55.6%	Claude Opus 5.518.6% · +18.6ppAnthropic GPT-6 Astra8.0% · +8.0ppOpenai Kimi K35.9% · +5.9ppMoonshotai
2	通用General	8.0%	Claude Opus 5.515.6% · +15.6ppAnthropic Gemini 3.8 Flash4.8% · +4.8ppGoogle Gemini 3.1 Pro Preview4.1% · +4.1ppGoogle
3	艺术与娱乐Arts & entertainment	6.8%	Claude Sonnet 4.510.5% · +10.5ppAnthropic Claude Opus 5.56.9% · +6.9ppAnthropic Gemini 3.8 Flash5.2% · +5.2ppGoogle
4	业务运营Business operations	5.5%	Claude Opus 5.511.5% · +11.5ppAnthropic GPT-5.6 Sol7.8% · +7.8ppOpenai GPT-6.1 Sol4.6% · +4.6ppOpenai
5	科学与工程Science & engineering	4.2%	Claude Opus 5.520.0% · +20.0ppAnthropic Kimi K316.4% · +16.4ppMoonshotai Claude Opus 55.4% · +5.4ppAnthropic
6	游戏Gaming	3.1%	Claude Opus 5.511.5% · +11.5ppAnthropic GPT-6 Astra9.1% · +9.1ppOpenai Grok 4.76.8% · +6.8ppX Ai
7	金融Finance	2.8%	Gemini 3 Flash Preview14.1% · +14.1ppGoogle Claude Opus 5.511.1% · +11.1ppAnthropic Kimi K35.7% · +5.7ppMoonshotai
8	健康与养生Health & wellness	2.1%	Claude Sonnet 4.615.9% · +15.9ppAnthropic Claude Opus 5.56.9% · +6.9ppAnthropic Claude Opus 4.84.8% · +4.8ppAnthropic
9	法律Legal	2.1%	Gemini 3.1 Pro Preview14.4% · +14.4ppGoogle Claude Sonnet 57.5% · +7.5ppAnthropic Claude Sonnet 5.57.5% · +7.5ppAnthropic
10	教育Education	2.1%	Claude Opus 5.59.8% · +9.8ppAnthropic GPT-5.6 Sol6.2% · +6.2ppOpenai GLM 5.3 Flash5.1% · +5.1ppZ Ai
11	客户服务Customer service	1.7%	Gemini 3 Flash Preview7.5% · +7.5ppGoogle Claude Opus 5.55.6% · +5.6ppAnthropic GPT-4.15.4% · +5.4ppOpenai
12	市场与广告Marketing & advertising	1.7%	Claude Opus 5.512.6% · +12.6ppAnthropic Claude Sonnet 57.1% · +7.1ppAnthropic Gemini 3.8 Flash5.7% · +5.7ppGoogle
13	销售与电商Sales & commerce	1.4%	Claude Opus 5.59.6% · +9.6ppAnthropic Claude Sonnet 4.68.3% · +8.3ppAnthropic Claude Sonnet 5.55.8% · +5.8ppAnthropic
14	政务与政治Government & politics	0.9%	GPT-6 Sol9.7% · +9.7ppOpenai Kimi K38.8% · +8.8ppMoonshotai Claude Opus 5.55.9% · +5.9ppAnthropic
15	情感与关系Relationships	0.8%	DeepSeek V4 Flash 042313.0% · +13.0ppDeepseek Claude Sonnet 4.512.8% · +12.8ppAnthropic GLM 5.26.8% · +6.8ppZ Ai
16	旅行与酒店Travel & hospitality	0.5%	Claude Opus 5.515.1% · +15.1ppAnthropic Gemini 3.8 Flash7.0% · +7.0ppGoogle GPT-6.1 Sol4.9% · +4.9ppOpenai
17	生活方式Lifestyle	0.4%	Claude Opus 5.512.1% · +12.1ppAnthropic Gemini 3.8 Flash10.3% · +10.3ppGoogle Claude Sonnet 58.9% · +8.9ppAnthropic

工具调用与图像处理

2026-10-26 当周

按周统计工具调用次数与图像处理张数

工具调用最多的模型

排名	模型	调用量	占比
1	DeepSeek V4.1 FlashDeepSeek	92.3M	19.4%
2	GLM 5.3 FlashZ.ai	45.5M	9.6%
3	DeepSeek V4.1 FlashDeepSeek	40.9M	8.6%
4	GPT-6 LunaOpenAI	36.3M	7.6%
5	Hy4 previewTencent	26.2M	5.5%
6	DeepSeek V4 Flash 0423DeepSeek	14.6M	3.1%
7	Claude Haiku 5.5Anthropic	12.9M	2.7%
8	Claude Opus 5.5Anthropic	11.9M	2.5%
9	GLM 5.3Z.ai	11.6M	2.4%

处理图像最多的模型

排名	模型	处理量	占比
1	GLM 5.3 FlashZ.ai	115.9M	13.5%
2	GPT-6 LunaOpenAI	111.9M	13.0%
3	DeepSeek V4.1 FlashDeepSeek	93.5M	10.9%
4	Gemini 3.1 Flash LiteGoogle	68.3M	7.9%
5	Jev 1.13TypeSafe	67.4M	7.8%
6	Gpt 6 Luna DecisionsOpenAI	40.2M	4.7%
7	DeepSeek V4.1 FlashDeepSeek	29.6M	3.4%
8	Muse Spark 1.3 ContributorMeta	27.3M	3.2%
9	Claude Opus 5.5Anthropic	25.4M	2.9%

生成图像最多的模型

排名	模型	生成量	占比
1	Nano Banana 2 (Gemini 3.1 Flash Image)Google	549.7K	17.1%
2	Gpt Image 2.5 SunburstOpenAI	398.4K	12.4%
3	Muse Image 1.0 EvalMeta	354.0K	11.0%
4	Gpt Image 2.5 FlareOpenAI	262.1K	8.2%
5	Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)Google	225.5K	7.0%
6	Gpt Image 2OpenAI	191.6K	6.0%
7	Nano Banana 2.1Google	163.2K	5.1%
8	Nano Banana (Gemini 2.5 Flash Image)Google	148.8K	4.6%
9	Nano Banana Pro (Gemini 3 Pro Image)Google	133.4K	4.2%