Claude Sonnet 5.5 vs. Grok 4.7：发布比较

Claude Sonnet 5.5 与 Grok 4.7 发布比较：分别包含 5 个和 3 个模型，各自具有不同的智能、性能与价格特征。下方对全部 8 个模型的关键指标进行了比较。

智能方面，Claude Sonnet 5.5 得分最高：Claude Sonnet 5.5 (Max, Default Fallback)（55），Grok 4.7 为 Grok 4.7 (Xhigh)（35）。
每任务耗时方面，Claude Sonnet 5.5 最快：Claude Sonnet 5.5 (Low, Default Fallback)（2.7 分钟），Grok 4.7 为 Grok 4.7 (Low)（9.9 分钟）。
单任务成本方面，Claude Sonnet 5.5 最低：Claude Sonnet 5.5 (Low, Default Fallback)（$0.51），Grok 4.7 为 Grok 4.7 (Low)（$2.92）。

比较的发布


	Claude Sonnet 5.5 (max)	Claude Sonnet 5.5 (xhigh)	Claude Sonnet 5.5 (high)	Grok 4.7 (xhigh)	Grok 4.7 (high)	Grok 4.7 (low)	Claude Sonnet 5.5 (medium)	Claude Sonnet 5.5 (low)
Intelligence Index	55	50	42	35	35	31	31	26
财务与会计	56	49	44	52	51	45	41	36
战略与运营	58	54	51	55	53	49	47	42
法律	52	47	44	47	47	44	39	37
医疗与健康	54	47	44	47	46	43	40	37
工程	58	54	46	42	42	38	37	33
经济学	54	48	45	46	46	41	39	35
AA-Briefcase v1.1	1813	1742	1634	1637	1629	1493	1439	1268
AutomationBench-AA	71.8%	65.5%	59.4%	65.6%	63.5%	56.7%	54.9%	49.4%
Terminal-Bench 4.0	63.6%	57.1%	43.9%	25.8%	24.7%	16.2%	29.8%	20.7%
FrontierCode 2.0	41.7%	35.6%	25.7%	28.7%	27.8%	21.8%	11.0%	7.1%
AA-Omniscience v1.1	30	21	19	31	30	27	19	20
GDP.pdf	25.8%	24.6%	25.2%	20.0%	23.2%	21.4%	20.2%	16.0%
HLE-Diamond	53.3%	44.1%	33.8%	23.2%	23.6%	18.5%	23.8%	20.6%
CritPt v2 (Hard Subset)	59.0%	57.5%	51.5%	33.5%	37.0%	43.5%	33.5%	26.0%
Terminal-Bench-Science 0.1	53.3%	52.4%	34.3%	11.4%	9.5%	4.8%	16.2%	10.5%

每任务成本美元	$10.02	$4.09	$1.64	$7.77	$6.18	$2.92	$0.78	$0.51
输入价格美元/100 万 Token	$2.00	$2.00	$2.00	$2.00	$2.00	$2.00	$2.00	$2.00
输出价格美元/100 万 Token	$10.00	$10.00	$10.00	$6.00	$6.00	$6.00	$10.00	$10.00
缓存命中价格美元/100 万 Token	$0.10	$0.10	$0.10	$0.50	$0.50	$0.50	$0.10	$0.10
缓存写入价格美元/100 万 Token	$2.50	$2.50	$2.50	--	--	--	$2.50	$2.50

每任务耗时分钟	31.1	15.2	7.0	24.8	20.4	9.9	3.6	2.7
输出速度tok/s	153	134	139	70	73	72	143	131
首个答案 Token 延迟秒	303.84	32.09	11.75	63.97	32.44	9.65	5.40	3.28
首 Token 延迟秒	303.84	32.09	11.75	63.97	32.44	9.65	5.40	3.28
端到端时间单次请求工作负载的平均值（秒）	604.99	93.83	52.69	152.80	89.97	38.27	31.82	26.90

上下文窗口	1M	1M	1M	500k	500k	500k	1M	1M
开发商	Anthropic	Anthropic	Anthropic	SpaceXAI	SpaceXAI	SpaceXAI	Anthropic	Anthropic
许可证	专有	专有	专有	专有	专有	专有	专有	专有
输入模态	支持：文本和图像	支持：文本和图像	支持：文本和图像	支持：文本和图像	支持：文本和图像	支持：文本和图像	支持：文本和图像	支持：文本和图像
输出模态	支持：文本	支持：文本	支持：文本	支持：文本	支持：文本	支持：文本	支持：文本	支持：文本
服务商

Updated

智能

Artificial Analysis Intelligence Index · Higher is better

每任务耗时

Inference time (minutes) per Intelligence Index task · Lower is better

每任务成本

Weighted average cost (USD) per Intelligence Index task · Lower is better

智能Updated

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v5.0.1 incorporates 9 evaluations: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), and Terminal-Bench-Science 0.1

No publicly available API

Artificial Analysis Intelligence Index v5.0.1 includes: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), Terminal-Bench-Science 0.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.

Intelligence Index 与每项任务的成本

Artificial Analysis Intelligence Index · Weighted average cost (USD) per Artificial Analysis Intelligence Index task

Most attractive quadrant

Claude Sonnet 5.5

Grok 4.7

Pareto line

AA-Briefcase v1.1

AutomationBench-AA

单任务成本（美元，对数刻度）

Weighted average cost per Intelligence Index task. Each evaluation’s cost is calculated from input, cache hit, cache write, reasoning, and answer token prices, divided by task count, and weighted by its Intelligence Index weight.

成本

每项 Intelligence Index 任务的成本

Weighted average cost (USD) per Intelligence Index task · Segmented by token type · Lower is better

No publicly available API

速度与延迟

Intelligence Index 每项任务耗时

Output tokens divided by end-to-end output speed (minutes per Intelligence Index task) · Speed at the model's own endpoint or the median across providers · Lower is better

No publicly available API

The weighted average time (minutes) per Artificial Analysis Intelligence Index task: output tokens per task divided by output speed, weighted by each benchmark's weight in the Intelligence Index.

Speed and latency metrics average each endpoint's runs over the past 5 days, and the error rate pools them. We run the benchmark every 6 hours, and endpoints of deprecated models every 18 hours.

能力得分

能力指数

衡量模型在特定能力与行业中的表现

Finance & Accounting Index

Incorporates 5 evaluations: AA-Briefcase v1.1, AA-Omniscience v1.1, HLE-Diamond, AutomationBench-AA, GDP.pdf · Higher is better

Strategy & Ops Index

Incorporates 4 evaluations: AutomationBench-AA, AA-Briefcase v1.1, AA-Omniscience v1.1, GDP.pdf · Higher is better

Legal Index

Incorporates 5 evaluations: AA-Omniscience v1.1, AA-Briefcase v1.1, HLE-Diamond, GDP.pdf, AutomationBench-AA · Higher is better

Healthcare & Medical Index

Incorporates 5 evaluations: AA-Omniscience v1.1, AA-Briefcase v1.1, MLCR-AA, AutomationBench-AA, HLE-Diamond · Higher is better

Engineering Index

Incorporates 7 evaluations: AA-Omniscience v1.1, CritPt v2 (Hard Subset), Terminal-Bench-Science 0.1, HLE-Diamond, Terminal-Bench 4.0, FrontierCode 2.0, AA-Briefcase v1.1 · Higher is better

Economics Index

Incorporates 4 evaluations: AA-Omniscience v1.1, AA-Briefcase v1.1, HLE-Diamond, GDP.pdf · Higher is better

基准测试

AA-PersonalityBenchNew

模型 Avs模型 B

AA-PersonalityBench：Expressiveness 指数，Claude Sonnet 5.5 (max) vs Grok 4.7 (xhigh)

根据两者的 Elo 差距，每个模型的回答在各特质上被判定为更显著的概率 · 50% 表示势均力敌

Claude Sonnet 5.5 (max)

Grok 4.7 (xhigh)

Eloquent

21%

79%

Natural Sounding

25%

75%

Metaphorical

72%

28%

Lexically Varied

64%

36%

Literate

48%

52%

Claude Sonnet 5.5 vs. Grok 4.7：发布比较

智能

每任务耗时

每任务成本

智能Updated

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index

Intelligence Index 与每项任务的成本

Cost per Intelligence Index Task

Artificial Analysis Intelligence Index

成本

每项 Intelligence Index 任务的成本

Cost per Intelligence Index Task

速度与延迟

Intelligence Index 每项任务耗时

Time per Intelligence Index Task

Time Period

能力得分

能力指数

基准测试

AA-PersonalityBenchNew

AA-PersonalityBench：Expressiveness 指数，Claude Sonnet 5.5 (max) vs Grok 4.7 (xhigh)