模型排行榜

#	模型	上下文	日 Tokens	日请求	日变化
148	L liquid/lfm-2.5-2.6b-20260811 free	—	13.7B	1.3M	27.2%
149	qwen/qwen3.6-flash	—	12.9B	2.1M	36.5%
150	qwen/qwen3.5-122b-a10b-20260224	—	12.9B	1.7M	29.5%
151	google/gemini-embedding-001	—	12.4B	15.1M	82.8%
152	openai/gpt-4o	—	12.4B	5.7M	15.2%
153	qwen/qwen3-vl-8b-instruct	—	12.3B	10.5M	7.5%
154	z-ai/glm-4.6	—	12.1B	870.5K	10.7%
155	deepseek/deepseek-r1-0528	—	11.9B	1.2M	7.5%
156	bytedance-seed/seed-2.0-mini-20260224	—	11.8B	2.3M	7.6%
157	ibm-granite/granite-4.0-h-micro	—	11.5B	3.1M	365.8%
158	google/gemma-3-12b-it	—	11.4B	8M	17.7%
159	openai/gpt-4o-mini-2024-07-18	—	11.2B	6.2M	83.2%
160	meta/muse-spark-1.1-20260709	—	10.8B	701.9K	55.6%
161	openai/gpt-5-2025-08-07	—	10.5B	1.9M	16.3%
162	V voyageai/rerank-2.5-20260727	—	10.2B	2.2M	21.6%
163	A aion-labs/aion-3.0-20260707	—	9.8B	480.8K	2.2%
164	qwen/qwen3-embedding-4b	—	9.5B	4.9M	45.0%
165	qwen/qwen3.5-27b-20260224	—	9.2B	2.8M	12.2%
166	qwen/qwen3-next-80b-a3b-instruct-2509	—	8.8B	1.8M	39.2%
167	qwen/qwen3-vl-30b-a3b-instruct	—	8.6B	7.2M	2.4%
168	meituan/longcat-2.0-20260720	—	7.8B	349.7K	35.7%
169	openai/gpt-oss-safeguard-20b	—	7.6B	3.3M	4.1%
170	deepseek/deepseek-v3.1-terminus	—	7.5B	838.1K	2.2%
171	nvidia/nemotron-3-super-120b-a12b-20230311	—	7.5B	981.3K	14.7%
172	nvidia/nemotron-3-nano-30b-a3b	—	7.4B	888K	124.8%
173	qwen/qwen3-reranker-8b	—	7.2B	600.8K	80.1%
174	qwen/qwen3-8b-04-28	—	7.1B	2.1M	31.6%
175	mistralai/mistral-large-2512	—	7B	2.5M	59.1%
176	tencent/hy3-preview-20260421	—	6.5B	953.4K	22.3%
177	mistralai/mistral-embed-2312	—	6.1B	1.1M	10.8%
178	poolside/laguna-xs-2.1-20260625	—	6B	256.6K	45.1%
179	google/gemini-3.7-flash-20260813 batch	—	5.8B	1.2M	58.1%
180	S sao10k/l3-lunaris-8b	—	5.7B	2.3M	16.9%
181	thinkingmachines/inkling-small-20260730	—	5.6B	1M	63.4%
182	z-ai/glm-4.5-air	—	5.5B	1.2M	29.7%
183	openai/gpt-5.6-luna-20260709 batch	—	5.4B	946.5K	50.1%
184	I inclusionai/ling-3.0-flash-vl-20260910 free	—	5.1B	112.7K	—
185	mistralai/ministral-8b-2512	—	5B	3.4M	5.5%
186	qwen/qwen-2.5-7b-instruct	—	4.9B	5.1M	3.4%
187	qwen/qwen3-coder-30b-a3b-instruct	—	4.7B	620.9K	13.5%
188	z-ai/glm-5v-turbo-20260401	—	4.5B	231.8K	22.9%
189	amazon/nova-lite-v1	—	4.2B	1.6M	73.1%
190	qwen/qwen3-coder-flash	—	4.1B	127.5K	147.4%
191	qwen/qwen3-14b-04-28	—	4.1B	1.7M	22.6%
192	google/gemini-3.8-flash-20260902 batch	—	3.9B	863.5K	620.1%
193	qwen/qwen3-max	—	3.8B	666K	68.2%
194	I inception/mercury-2-20260304	—	3.8B	1M	37.7%
195	google/gemini-3.1-flash-image-20260528	—	3.6B	1.3M	12.5%

口径：日消耗 = prompt + completion + reasoning tokens；变化为相对前一日的比例。页面尚未公开来源链接，数据仅供参考。