MB / MODELS

模型排行榜

按日 Token 消耗排序，覆盖 546 个在榜模型

数据日期 2026-09-10 快照数据最近同步 14:58 回到首页

完整榜单 共 546 个模型 · 当前第 8/12 页

#	模型	上下文	日 Tokens	日请求	日变化
340	bytedance/ui-tars-1.5-7b	—	235.9M	110.7K	7.9%
341	qwen/qwen3-vl-235b-a22b-thinking	—	232.5M	50.2K	26.7%
342	krea/krea-2-medium-20260720	—	210.8M	48.8K	76.8%
343	tencent/hunyuan-a13b-instruct	—	187.4M	41.3K	126.7%
344	meta-llama/llama-3.2-1b-instruct	—	172.8M	149.3K	71.8%
345	cohere/command-r-08-2024	—	169.2M	101.4K	15.2%
346	google/lyria-3-pro-preview-20260330	—	164.5M	42.3K	695.1%
347	upstage/solar-pro-3	—	163.2M	93.8K	199.2%
348	thedrummer/skyfall-36b-v2	—	160.9M	43K	8.2%
349	qwen/qwen3-vl-30b-a3b-thinking	—	149.2M	42.9K	33.1%
350	mistralai/mistral-saba-2502	—	147.2M	46.7K	40.7%
351	mistralai/mixtral-8x22b-instruct	—	136.1M	83.8K	2.7%
352	mistralai/voxtral-small-24b-2507	—	133.5M	210.3K	12.9%
353	qwen/qwen3-30b-a3b-thinking-2507	—	129.7M	32.1K	75.4%
354	liquid/lfm-2.5-embedding-350m-20260818 free	—	128M	97.2K	46.9%
355	perplexity/sonar-reasoning-pro	—	123.1M	61.3K	16.9%
356	openai/gpt-6-astra-20260903 batch	—	122.7M	59K	—
357	google/gemini-3.5-flash-20260519 batch	—	121.5M	13K	46.5%
358	qwen/qwen-image-3-20260805	—	120M	27.4K	7.1%
359	openai/o4-mini-high-2025-04-16	—	118M	16.7K	12.8%
360	openai/gpt-5-image	—	117.6M	27K	31.2%
361	qwen/qwen3-next-80b-a3b-thinking-2509	—	115.8M	19.7K	56.8%
362	openai/gpt-5.6-sol-pro-20260709 batch	—	115.7M	2.4K	230.3%
363	undi95/remm-slerp-l2-13b	—	114.4M	60K	16.8%
364	intfloat/e5-large-v2-20251117	—	107.7M	204.3K	4.9%
365	perplexity/sonar-deep-research	—	103.4M	6K	5.0%
366	cohere/north-mini-code-20260617	—	102.8M	4.9K	454.9%
367	anthracite-org/magnum-v4-72b	—	100.4M	31.5K	20.3%
368	z-ai/glm-5.3-20260816 batch	—	97.7M	7.1K	—
369	hexgrad/kokoro-82m	—	95.3M	582.8K	10.6%
370	google/gemma-2-27b-it	—	94.9M	152K	125.6%
371	google/gemini-embedding-2 batch	—	94.9M	1.7M	72.5%
372	nvidia/nemotron-3-nano-omni-30b-a3b-reasoning-20260428	—	94M	4.7K	21.2%
373	openai/gpt-image-2.5-flare-20260908	—	93.3M	38.7K	—
374	deepseek/deepseek-r1-distill-llama-70b	—	93.1M	37.8K	0.6%
375	black-forest-labs/flux.2-max	—	88.9M	9.2K	75.5%
376	openai/gpt-4-turbo	—	88M	114.5K	64.7%
377	openai/gpt-6-astra-pro-20260903 batch	—	86.1M	1.5K	—
378	amazon/nova-pro-v1	—	77.2M	26.5K	24.1%
379	amazon/nova-premier-v1	—	76.8M	30.6K	2.4%
380	morph/morph-v3-fast	—	72.2M	7K	46.0%
381	google/gemini-3.1-pro-preview-20260219 batch	—	69.9M	13.5K	13.4%
382	baai/bge-large-en-v1.5-20251117	—	64.3M	76.4K	15.2%
383	cohere/command-a-03-2025	—	62.4M	45.7K	53.3%
384	morph/morph-v3-large	—	62.2M	8.3K	18.8%
385	krea/krea-2-medium-turbo-20260720	—	61.9M	13.9K	49.1%
386	sentence-transformers/paraphrase-minilm-l6-v2-20251117	—	59.4M	12.1K	160.3%
387	qwen/qwen-image-3-pro-20260805	—	58.8M	12.9K	1.8%

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。