模型排行榜

#	模型	上下文	日 Tokens	日请求	日变化
196	V voyageai/voyage-4-20260727	—	3.5B	2.2M	9.5%
197	x-ai/grok-build-0.1-20260520	—	3.4B	105.2K	13.2%
198	I inception/mercury-2-20260304	—	3.4B	867K	8.7%
199	bytedance-seed/seed-1.6-flash-20250625	—	3.2B	2.7M	21.9%
200	meta-llama/llama-3.1-70b-instruct	—	3.1B	2M	4.0%
201	google/gemini-embedding-2-preview	—	3.1B	4M	8.3%
202	mistralai/mistral-medium-3.5-20260430	—	3.1B	947.2K	51.4%
203	openai/gpt-4o-2024-11-20	—	3B	659.9K	0.7%
204	bytedance-seed/seed-2-1-turbo-20260810	—	3B	301.7K	103.3%
205	qwen/qwen3.5-plus-20260216	—	3B	230K	3.6%
206	google/gemini-2.5-flash-lite batch	—	3B	1.4M	72.9%
207	qwen/qwen3-14b-04-28	—	2.9B	1.1M	48.7%
208	qwen/qwen2.5-vl-72b-instruct	—	2.8B	2.2M	0.0%
209	I inclusionai/ling-3.0-flash-fin-20260827	—	2.8B	167.1K	193.3%
210	V voyageai/voyage-4-large-20260727	—	2.8B	913.8K	0.9%
211	openai/o4-mini-2025-04-16	—	2.7B	521.6K	2.1%
212	mistralai/ministral-14b-2512	—	2.7B	1.1M	14.5%
213	anthropic/claude-3-haiku	—	2.7B	1.1M	22.2%
214	moonshotai/kimi-k2-thinking-20251106	—	2.7B	463.1K	8.1%
215	deepseek/deepseek-r1	—	2.7B	379.4K	5.9%
216	openai/gpt-5.2-codex-20260114	—	2.6B	152.2K	27.5%
217	z-ai/glm-5-turbo-20260315	—	2.5B	129.6K	35.4%
218	qwen/qwen-2.5-72b-instruct	—	2.4B	2.6M	19.6%
219	mistralai/mistral-medium-3.1	—	2.4B	641.1K	2.9%
220	google/gemini-3.1-flash-lite-image-20260630	—	2.4B	943.7K	21.7%
221	perplexity/sonar	—	2.4B	2.5M	9.6%
222	moonshotai/kimi-k2	—	2.4B	603.6K	7.9%
223	qwen/qwen3.5-plus-20260420	—	2.3B	169.8K	56.0%
224	V voyageai/voyage-4-lite-20260727	—	2.3B	1.2M	176.4%
225	A aion-labs/aion-3.0-mini-20260707	—	2.2B	157.1K	1.2%
226	z-ai/glm-4.6-20251208	—	2.2B	543.6K	43.6%
227	nvidia/llama-nemotron-embed-vl-1b-v2-20260224 free	—	2.2B	359.2K	111.0%
228	amazon/nova-2-lite-v1	—	2.2B	751.6K	18.3%
229	bytedance-seed/seedream-4.5-20251203	—	2.2B	101.6K	0.1%
230	tencent/hy-mt2-30b-a3b-20260521	—	2.2B	53.8M	399.5%
231	qwen/qwen3-coder-plus	—	2.2B	76K	28.2%
232	mistralai/mistral-medium-3	—	2.1B	145.6K	380.8%
233	nvidia/llama-nemotron-rerank-vl-1b-v2 free	—	2.1B	94.4K	12.8%
234	openai/gpt-5.6-luna-pro-20260709 batch	—	2.1B	49.1K	84.1%
235	openai/gpt-image-2	—	2B	1.3M	22.1%
236	google/gemini-3.1-pro-preview-customtools-20260219	—	2B	44.6K	87.2%
237	google/gemini-3.6-flash-20260721 batch	—	2B	941.5K	42.5%
238	meta-llama/llama-guard-4-12b	—	2B	2.4M	1.8%
239	stepfun/step-3.5-flash	—	2B	119K	1.9%
240	qwen/qwen3-235b-a22b-04-28	—	1.9B	538.1K	39.3%
241	ibm-granite/granite-4.2-8b-20260831	—	1.9B	406.4K	7.7%
242	qwen/qwen3-30b-a3b-04-28	—	1.9B	666.4K	32.9%
243	openai/gpt-chat-latest-20260505	—	1.8B	196.5K	23.5%

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。