LynxModel

AI Model Intelligence RankingsAI 模型智能排行

62 frontier models, one transparent 100-point Intelligence Score across 11 public benchmarks (reasoning, coding, knowledge, math, agent, human preference). Every score links to its source. Data refreshed 2026-09-20. Click any column to sort.

62 个前沿模型,一套透明的百分制智能评分,覆盖 11 个公开 benchmark(推理/代码/知识/数学/智能体/人类偏好)。每个分数都带来源链接。数据更新于 2026-09-20。点表头可排序。

Find my model帮我选型 Pricing & value价格与性价比 Head-to-head模型对战 Image arena图片生成榜 Video arena视频生成榜 How the score works评分方法 Benchmark explorerBenchmark 百科

Rea = Reasoning推理 Cod = Coding代码 Kno = Knowledge知识 Mat = Math数学 Agt = Agent智能体 Pre = Preference偏好

⚖ Adjust the weights — the ranking recomputes live⚖ 调权重——排名实时重算

Reasoning推理30

Coding代码25

Knowledge知识15

Math数学15

Agent智能体10

Preference偏好5

Same formula as the published score: weighted average over the dimensions a model actually has, renormalized by the weights present. Below 30% coverage a model shows no score; below 60% it carries a low-data badge.与公开评分同一公式:按模型实际拥有的维度加权平均、按出席权重重归一;覆盖<30% 不出分,<60% 打"数据少"徽标。

Intelligence vs price — the value map智能分 vs 价格——性价比地图

Each dot is a scored model: higher = smarter, further left = cheaper. Top-left is the sweet spot. Color = provider, dashed lines = medians.每个点是一个有评分的模型:越往上越聪明,越往左越便宜,左上角是甜区。颜色=厂商,虚线=中位数。 Click to enlarge. 点击放大。

Intelligence vs price scatter — each dot a scored model, color = provider

#	Model模型	Score总分	Rea推理	Cod代码	Kno知识	Mat数学	Agt智能体	Pre偏好	Ctx上下文	$ in/1M输入$/1M	$ out/1M输出$/1M
1	GPT-5.6 Sol(max)OpenAI	92.4	90.4	96.3	96.2	96	73.8	100	1.1M	$2.00	$10
2	Claude Fable 5(max)Anthropic	92.3	95.7	96.2	98.8	96.9	40.3	70.8	1M	$10	$50
3	DeepSeek V4 Flash 0731DeepSeek	91.1	87.2	94.3	95.1	93.1	—	—	1.3M	$0.040	$0.160
4	GPT-5.6 Terra(max)OpenAI	89.7	88.8	88.8	93.6	93.9	87.6	74.8	1.1M	$2.00	$12
5	GPT-5.6 Sol Pro(max)OpenAI	88.8 low data数据少	91.9	—	90.7	—	—	64.5	1.1M	$2.00	$10
6	DeepSeek V4 Pro 0423(max)DeepSeek	87.4	89.2	85.9	94.5	85.5	97.7	58.8	1M	$0.422	$0.845
7	Gemini 3.1 Pro Preview(high)Google	86.1	90.5	84.6	100	96.3	50.6	65.8	1M	$2.00	$12
8	GLM 5.2(max)Z.ai	85.4	95.4	73	96	99.3	37.7	62	1M	$0.650	$2.04
9	Hy3Tencent	85.3 low data数据少	88.3	—	—	—	94.1	58.5	262K	$0.083	$0.330
10	Qwen3.7 Max(on/off开/关)Qwen	85.2	95	85.7	96.8	57.2	96.1	62.8	1M	$1.48	$4.42
11	Grok 4.20(high)xAI	85.1	91.8	75.8	93.2	82.2	98	63	2M	$1.25	$2.50
12	Kimi K3(always-on常开)Moonshot AI	85	87.5	97.7	95	70.7	37.7	66.5	1M	$1.70	$8.50
13	Qwen3.8 Max (0902)Qwen	84.9	97.4	93.2	95.7	54.9	64.2	67	1M	$2.00	$6.00
14	GPT-5.6 Luna(max)OpenAI	84.7	80.1	86.3	92.9	92.4	—	57.2	1.1M	$0.200	$1.20
15	Claude Opus 4.8(max)Anthropic	84.2	89.4	90	96.8	75.9	58.8	62.8	1M	$5.00	$25
16	Gemini 3.5 Flash(high)Google	83.6	83.3	85.1	96.7	70.9	96.8	63.5	1M	$1.50	$9.00
17	DeepSeek V4 Flash 0423(max)DeepSeek	83.5	73.7	88.4	93.1	97	97.1	53	1M	$0.036	$0.071
18	Qwen3.7 Plus(on/off开/关)Qwen	83.4	77.5	86.9	95.6	85.7	94.4	58.8	1M	$0.320	$1.28
19	Claude Opus 5(max)Anthropic	83.3	96.4	85.9	88.5	67.2	62	66.5	1M	$5.00	$25
20	Muse Spark 1.2meta	83 low data数据少	82.7	87.5	—	—	—	68.8	1M	$1.25	$4.25
21	GPT-5.5(max)OpenAI	82	80.5	89.1	93	85.5	55.5	64.8	1.1M	$5.00	$30
22	Qwen3.7 Flash(on/off开/关)Qwen	82	78.5	87.3	96.4	67.3	—	63	1M	$0.030	$0.130
23	MiniMax M3MiniMax	81.8	80.1	82	90.9	83.9	82.4	55.2	1M	$0.300	$1.20
24	Step 3.7 FlashStepFun	81.4	86.4	74.3	—	97.3	100	43	262K	$0.200	$1.15
25	Claude Sonnet 5(high)Anthropic	80.9	89.1	88.2	94.5	49.1	75.8	59.7	1M	$2.00	$10
26	Kimi K2.7 CodeMoonshot AI	80.2	69.3	82	91.4	100	91.5	59.5	262K	$0.706	$3.21
27	Gemini 3.6 Flash(high)Google	80.1	84.5	91.7	59.8	77.7	—	65.3	1M	$0.750	$3.75
28	Grok 4.5(high)xAI	77.9	83.7	88.6	96.3	58.7	43.3	61.5	500K	$2.00	$6.00
29	Seed-2.0-MiniByteDance Seed	77.7 low data数据少	82.7	69.8	—	87	—	—	262K	$0.100	$0.400
30	MiMo-V2.5-ProXiaomi	77.2 low data数据少	87.4	—	—	37.3	95.6	61.3	1.1M	$0.435	$0.870
31	Nemotron 3 UltraNVIDIA	76.9	68.7	81.5	93.7	—	84.6	51	262K	$0.600	$2.40
32	MiniMax M2.7MiniMax	75.8	71.1	77.7	88.6	78.3	77.3	48.3	205K	$0.300	$1.20
33	Grok 4.3(high)xAI	74.3	77.7	70.8	92.7	53.4	99.2	54.7	1M	$1.25	$2.50
34	GPT-5.6 Luna Pro(max)OpenAI	71.5 low data数据少	85.4	22	—	—	—	56.7	1.1M	$0.200	$1.20
35	ERNIE 4.5 VL 424B A47B Baidu	71.2	68	49.9	83.8	78	—	—	123K	$0.420	$1.25
36	Kimi K2 Thinking(always-on常开)Moonshot AI	71.1	65.5	69.8	94.6	61.9	79.1	51.8	262K	$0.600	$2.50
37	DeepSeek V3.2(on/off开/关)DeepSeek	70.4	65.3	75.4	91.8	50.6	81.1	50.5	164K	$0.269	$0.400
38	GLM 5.1(on/off开/关)Z.ai	70.2	66.2	66.9	78.5	67.4	86.5	61	205K	$0.966	$3.04
39	Muse Spark 1.1meta	69.5	86.1	82.9	—	39.6	58	66.5	1M	$1.25	$4.25
40	Mistral Medium 3.5Mistral AI	65.5	42.9	69.4	87.8	86.3	95.6	51	262K	$1.50	$7.50
41	Seed-2.0-LiteByteDance Seed	65.4	66.9	72.9	—	50.6	95.6	32.2	262K	$0.250	$2.00
42	gpt-oss-120bOpenAI	65	59	56.3	85.3	93.4	66.8	35.5	131K	$0.150	$0.600
43	Gemini 3.5 Flash Lite(high)Google	63.4	60.4	57.8	92.7	36	—	58.8	1M	$0.300	$2.50
44	Mercury 2Inception	58.7 low data数据少	53.8	28.8	—	91.1	71.9	—	128K	$0.250	$0.750
45	Gemini 3.1 Flash Lite(high)Google	57	60.5	49	89.6	34.3	31.8	52.2	1M	$0.250	$1.50
46	Claude Haiku 4.5(high)Anthropic	55	55.8	66.2	70.2	21.8	55.5	47.5	200K	$1.00	$5.00
47	Sonar ProPerplexity	52.5 low data数据少	44.3	—	81.5	29	—	—	200K	$3.00	$15
48	Command ACohere	50.5	36.5	31.4	76.9	90	81.9	32.8	256K	$2.50	$10
49	Mistral Small 4Mistral AI	50.2	31.4	58.7	57.1	93	41.8	—	262K	$0.150	$0.600
50	Mistral LargeMistral AI	49.9	26.6	52.8	79	63.4	—	48	128K	$2.00	$6.00
51	Command R+ (08-2024)Cohere	42.8 low data数据少	33	—	81.7	6.7	—	—	128K	$2.50	$10
52	Llama 4 MaverickMeta	42.5	41.5	39.6	86.9	11	18.1	48.5	1M	$0.188	$0.652
53	Nova Pro 1.0Amazon	37 low data数据少	30	—	74.6	7	14.2	—	300K	$0.800	$3.20
54	Phi 4Microsoft	35.5	32.8	—	77.1	6.7	0	8.3	16K	$0.070	$0.140
55	Llama 4 ScoutMeta	34.5	33.7	23.5	80.2	3.6	15.7	41.7	1.3M	$0.100	$0.300
56	Nova Premier 1.0Amazon	31.4 low data数据少	33.4	33.9	—	17	38.9	—	1M	$2.50	$13
—	GPT-5.6 Terra Pro(max)OpenAI	— low data数据少	72.8	77	—	—	—	60.2	1.1M	$2.00	$12
—	LongCat 2.0Meituan	— low data数据少	81.7	—	—	100	—	—	1M	$0.300	$1.20
—	Hunyuan A13B InstructTencent	— low data数据少	94.7	—	—	76.8	—	—	131K	$0.140	$0.570
—	Solar Pro 3Upstage	— low data数据少	75.8	—	—	84.3	87.6	—	131K	$0.150	$0.600
—	Palmyra X5Writer	— low data数据少 vendor data厂商数据	49.4	—	—	—	—	—	1M	$0.600	$6.00
—	KAT-Coder-Pro V2.5Kwaipilot	— low data数据少 vendor data厂商数据	—	—	—	72.5	—	—	262K	$0.740	$2.96

56 of 62 models have enough data (≥30% benchmark coverage) for a public score. Prices are API list prices per 1M tokens from OpenRouter. "vendor data" means >40% of available scores are vendor self-reported — treat with care. The Chinese board lists models reachable via CN-direct endpoints, ranked by score.56/62 个模型数据充足(benchmark 覆盖 ≥30%)予以出分。价格为 OpenRouter 每百万 token API 标价。"厂商数据" 表示可用成绩中厂商自报占比超 40%,请谨慎参考。中文榜 = 支持国内直连的模型按分数排序。