模型排行榜

#	模型	上下文	日 Tokens	日请求	日变化
100	openai/gpt-5.6-luna-pro-20260709	—	72.7B	1.1M	6.0%
101	meta/muse-spark-1.3-20260902	—	72.4B	1.5M	49.6%
102	qwen/qwen3.7-plus-20260602	—	68.1B	6.5M	19.9%
103	x-ai/grok-4.5-20260708	—	67.3B	5.5M	10.1%
104	anthropic/claude-4.5-sonnet-20250929	—	64.9B	6.8M	70.2%
105	stepfun/step-3.7-flash-20260528	—	62.3B	1.4M	88.2%
106	openai/gpt-6-luna-20260922 batch	—	55.2B	9.9M	27.9%
107	qwen/qwen3-235b-a22b-07-25	—	54.8B	23.6M	12.9%
108	L liquid/d1-20260930	—	54B	18.2M	22.4%
109	deepseek/deepseek-chat-v3.1	—	51.9B	18.2M	3.4%
110	perplexity/pplx-embed-v1-0.6B	—	51.2B	14.7M	258.2%
111	x-ai/grok-4.6-20260810	—	51B	1.6M	14.6%
112	L liquid/d1-ultra-20261015	—	50.4B	21.2M	5.9%
113	openai/gpt-5.4-20260305	—	50B	5.7M	26.2%
114	moonshotai/kimi-k2.7-code-20260612	—	49.6B	1.1M	14.6%
115	google/gemini-3.5-flash-20260519	—	45.8B	6M	6.8%
116	openai/gpt-4.1-2025-04-14	—	45.3B	13.3M	4.6%
117	openai/gpt-6.1-sol-pro-20260929	—	44B	358.6K	14.3%
118	google/gemini-4-carbon-20261110	—	43.3B	704K	—
119	x-ai/grok-4.3-20260430	—	42.9B	7.2M	5.3%
120	openai/gpt-5.2-20251211	—	40.7B	3.6M	9.4%
121	qwen/qwen3.5-35b-a3b-20260224	—	40.3B	10M	8.8%
122	qwen/qwen3.8-omni-flash-20260918	—	39.3B	4.6M	14.6%
123	qwen/qwen3.5-flash-20260224	—	38.8B	19.8M	1.5%
124	openai/gpt-6.1-terra-pro-20261026	—	38.7B	387.7K	17.4%
125	z-ai/glm-5-20260211	—	37.9B	2.3M	1.0%
126	microsoft/microsoft-decision-1-20261009	—	37.3B	13M	27.6%
127	z-ai/glm-5.3-flashx-legal-20261102	—	37B	786.8K	268.2%
128	baai/bge-m3-20251117	—	36.7B	20.8M	15.3%
129	openai/gpt-5-nano-2025-08-07	—	36.5B	12M	12.3%
130	z-ai/glm-4.7-20251222	—	34.9B	2.5M	15.6%
131	meta-llama/llama-3.1-8b-instruct	—	34.9B	37.9M	17.1%
132	openai/gpt-5.5-20260423	—	32.3B	4.5M	0.6%
133	openai/gpt-5.1-20251113	—	32.3B	3.5M	0.7%
134	N nace-ai/drex-max-20261016	—	31.7B	41.9M	52.5%
135	qwen/qwen3-coder-next-2025-02-03	—	30.6B	977.5K	1.4%
136	nvidia/nemotron-3-ultra-550b-a55b-20260604	—	30.5B	3.4M	17.3%
137	google/gemma-3-27b-it	—	29.6B	13.4M	32.2%
138	z-ai/glm-5.3-prime-20260921	—	29.6B	707K	9.4%
139	google/gemini-2.5-pro	—	29B	2.9M	5.8%
140	meta-llama/llama-3.3-70b-instruct	—	28.8B	19.6M	3.9%
141	qwen/qwen3-vl-235b-a22b-instruct	—	28.5B	8M	2.8%
142	V voyageai/rerank-2.5-20260727	—	26.9B	4.8M	16.1%
143	meta/muse-glimmer-30b-20260810	—	26.4B	1.7M	100.3%
144	z-ai/glm-5.1-20260406	—	26.3B	1.6M	1.4%
145	qwen/qwen3.8-max-prime-20260923	—	26.3B	427.7K	2.2%
146	mistralai/mistral-small-3.2-24b-instruct-2506	—	26.2B	11.2M	5.6%
147	qwen/qwen3-32b-04-28	—	25.7B	14.5M	16.3%

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。