Kimi K3 vs. GPT-5.6 Terra：发布比较

Kimi K3 与 GPT-5.6 Terra 发布比较：分别包含 2 个和 6 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，Kimi K3 得分最高：Kimi K3 (Max)（44），GPT-5.6 Terra 为 GPT-5.6 Terra (Max)（42）。
输出速度方面，GPT-5.6 Terra 最快：GPT-5.6 Terra (Max)（123 token/秒），Kimi K3 为 Kimi K3 (Low)（45 token/秒）。
单任务成本方面，GPT-5.6 Terra 最低：GPT-5.6 Terra (Non-reasoning)（$0.14），Kimi K3 为 Kimi K3 (Low)（$1.15）。

比较的发布


	Kimi K3 (max)	GPT-5.6 Terra (max)	GPT-5.6 Terra (xhigh)	GPT-5.6 Terra (high)	GPT-5.6 Terra (medium)	Kimi K3 (low)	GPT-5.6 Terra (low)	GPT-5.6 Terra (non-reasoning)
Intelligence Index	44	42	38	34	30	30	27	21
财务与会计	47	41	40	37	32	32	30	21
战略与运营	50	47	44	40	34	39	31	23
法律	49	41	40	37	34	34	31	22
医疗与健康	45	40	38	34	31	--	27	--
工程	43	43	38	34	31	30	28	21
经济学	53	45	44	42	37	37	34	24
AA-Briefcase v1.1	1501	1334	1333	1202	1039	986	1014	947
GDPval-AA v2.1	1533	1451	1447	1376	1264	1126	1108	1097
AutomationBench-AA	58.3%	59.6%	47.1%	42.0%	33.5%	49.3%	29.1%	12.5%
Terminal-Bench 4.0	12.6%	35.4%	10.1%	1.5%	1.0%	12.6%	1.5%	0.5%
SciCode	59.5%	55.0%	52.3%	52.4%	50.5%	52.7%	49.9%	45.1%
Humanity's Last Exam	46.9%	42.9%	41.9%	38.5%	33.3%	25.0%	29.2%	11.4%
GDP.pdf	22.0%	24.0%	24.6%	20.8%	17.0%	18.6%	17.6%	11.2%
CritPt	23.4%	30.0%	27.1%	22.9%	17.4%	3.1%	9.4%	2.0%
AA-Omniscience	20	0	−3	−3	−5	4	−7	−23
AA-LCR v1.1	88.7%	83.0%	79.0%	77.7%	74.0%	79.3%	71.3%	58.7%

每任务成本美元	$2.00	$1.40	$0.63	$0.34	$0.18	$1.15	$0.14	$0.14
输入价格美元/100 万 Token	$3.00	$2.00	$2.00	$2.00	$2.00	$3.00	$2.00	$2.00
输出价格美元/100 万 Token	$15.00	$12.00	$12.00	$12.00	$12.00	$15.00	$12.00	$12.00
缓存命中价格美元/100 万 Token	$0.30	$0.20	$0.20	$0.20	$0.20	$0.30	$0.20	$0.20
缓存写入价格美元/100 万 Token	--	$2.50	$2.50	$2.50	$2.50	--	$2.50	$2.50

输出速度Token/秒	42	123	110	102	103	45	100	100
首 Token 延迟秒	3.23	149.29	37.76	7.05	1.65	3.27	1.45	0.69
首个回答 Token 时间秒	51.25	149.29	37.76	7.05	1.65	48.20	1.45	0.69
端到端响应时间秒	63.26	153.35	42.29	11.95	6.53	59.43	6.45	5.67
每任务耗时秒	1,074.21	307.79	179.96	109.99	54.09	268.79	37.94	31.13

上下文窗口	1M	1M	1M	1M	1M	1M	1M	1M
开发商	Kimi	OpenAI	OpenAI	OpenAI	OpenAI	Kimi	OpenAI	OpenAI
许可证	开放	专有	专有	专有	专有	开放	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	+10					+2

智能

Artificial Analysis Intelligence Index · Higher is better

速度

Output tokens per second · Higher is better

每任务成本

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

智能

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1

Artificial Analysis Intelligence Index v4.3.2 includes: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1. 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

Kimi K3

GPT-5.6 Terra

Pareto line

AA-Briefcase v1.1

GDPval-AA v2.1

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 Artificial Analysis Intelligence Index task, segmented by token type. Lower is better

速度与延迟

输出速度

Output tokens per second · Higher is better

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Figures represent performance of the model's first-party API or the median across providers where a first-party API is not available.

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 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better

Strategy & Ops Index

Incorporates 6 evaluations: GDPval-AA v2.1, AA-Briefcase v1.1, AA-Omniscience, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better

Legal Index

Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AA-LCR v1.1, GDP.pdf, AutomationBench-AA · Higher is better

Healthcare & Medical Index

Incorporates 6 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, MLCR-AA, Humanity's Last Exam, AutomationBench-AA · Higher is better

Engineering Index

Incorporates 6 evaluations: AA-Omniscience, Humanity's Last Exam, CritPt, GDPval-AA v2.1, AA-Briefcase v1.1, Terminal-Bench 4.0 · Higher is better

Economics Index

Incorporates 5 evaluations: AA-Omniscience, Humanity's Last Exam, GDPval-AA v2.1, AA-Briefcase v1.1, AA-LCR v1.1 · Higher is better

Kimi K3 vs. GPT-5.6 Terra：发布比较

智能

速度

每任务成本

智能

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

输出速度

Output Speed

Model Performance Representation

能力得分

能力指数