模型排行榜

#	模型	上下文	日 Tokens	日请求	日变化
196	mistralai/ministral-3b-2512	—	3.5B	2.4M	44.9%
197	meta-llama/llama-3.1-70b-instruct	—	3.4B	2M	13.4%
198	bytedance-seed/seed-1.6-flash-20250625	—	3.3B	3.1M	70.7%
199	moonshotai/kimi-k2-0905	—	3.3B	1M	24.7%
200	x-ai/grok-build-0.1-20260520	—	3.2B	105.9K	22.3%
201	qwen/qwen2.5-vl-72b-instruct	—	3.2B	2.4M	23.3%
202	anthropic/claude-3-haiku	—	3.2B	1.1M	7.3%
203	google/gemini-embedding-2-preview	—	3.1B	3.8M	5.1%
204	V voyageai/voyage-4-large-20260727	—	3B	1M	29.9%
205	openai/gpt-4o-2024-11-20	—	2.9B	399.2K	2.4%
206	openai/o4-mini-2025-04-16	—	2.9B	522K	10.7%
207	moonshotai/kimi-k2-thinking-20251106	—	2.8B	360.7K	10.9%
208	z-ai/glm-5-turbo-20260315	—	2.8B	136.3K	33.7%
209	google/gemini-3.1-flash-lite-image-20260630	—	2.8B	1.1M	4.7%
210	I inclusionai/ling-3.0-flash-fin-20260827	—	2.7B	233.2K	544.0%
211	mistralai/ministral-14b-2512	—	2.7B	1.1M	19.6%
212	openai/gpt-5.2-codex-20260114	—	2.7B	128.9K	101.6%
213	qwen/qwen3.5-plus-20260216	—	2.7B	220.3K	1.3%
214	deepseek/deepseek-r1	—	2.6B	412.6K	2.6%
215	qwen/qwen-2.5-72b-instruct	—	2.5B	2.1M	13.4%
216	mistralai/mistral-medium-3.5-20260430	—	2.5B	1M	69.4%
217	google/gemini-3.1-pro-preview-customtools-20260219	—	2.5B	50.2K	267.9%
218	openai/gpt-5.6-luna-pro-20260709 batch	—	2.4B	81.3K	689.7%
219	google/gemini-3.6-flash-20260721 batch	—	2.4B	1.1M	13.2%
220	A aion-labs/aion-3.0-mini-20260707	—	2.4B	159.8K	13.2%
221	moonshotai/kimi-k2	—	2.4B	529.5K	28.9%
222	perplexity/sonar	—	2.3B	2.5M	1.6%
223	V voyageai/voyage-4-20260727	—	2.3B	2.1M	32.0%
224	mistralai/mistral-medium-3.1	—	2.2B	586.8K	10.1%
225	bytedance-seed/seedream-4.5-20251203	—	2.2B	100.2K	1.5%
226	openai/gpt-image-2	—	2.1B	1.4M	39.0%
227	qwen/qwen3-30b-a3b-04-28	—	2.1B	645.2K	25.3%
228	qwen/qwen3.5-plus-20260420	—	2.1B	183.4K	31.2%
229	bytedance-seed/seed-2-1-turbo-20260810	—	2B	149.2K	63.6%
230	nvidia/llama-nemotron-rerank-vl-1b-v2 free	—	2B	94.7K	14.7%
231	google/gemini-2.5-flash-lite batch	—	2B	857.6K	55.9%
232	meta-llama/llama-guard-4-12b	—	2B	2.4M	1.3%
233	stepfun/step-3.5-flash	—	1.9B	116K	3.6%
234	qwen/qwen3-235b-a22b-04-28	—	1.9B	508.8K	22.1%
235	amazon/nova-2-lite-v1	—	1.9B	666.1K	11.6%
236	qwen/qwen3-coder-plus	—	1.8B	59.8K	4.5%
237	qwen/qwen3-235b-a22b-thinking-2507	—	1.8B	195.2K	18.0%
238	ibm-granite/granite-4.2-8b-20260831	—	1.7B	381.8K	4.2%
239	openai/gpt-5.4-image-2-20260421	—	1.7B	227.5K	1.8%
240	nvidia/nemotron-3.5-content-safety-20260604 free	—	1.7B	642.7K	5.8%
241	openai/gpt-chat-latest-20260505	—	1.6B	193.3K	12.1%
242	google/gemini-2.5-flash-image	—	1.6B	929.8K	11.2%
243	openai/gpt-5.1-codex-mini-20251113	—	1.6B	70.2K	27.9%

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。