排行榜/Arena 人类盲测/Code 编程/内容创作工具

Arena Code 编程 · 内容创作工具

前端 / 全栈代码生成盲测，含 11 个细分场景子榜 · 完整榜单 138 个模型

Overall 总榜143 Fullstack 全栈82 Frontend 前端143 HTML142 React127 品牌营销站142 参考图还原142 数据分析应用142 消费级产品142 游戏142 仿真模拟142 内容创作工具138

Rank	区间	Model	厂商	License	Elo 分	投票数	输入 $/M	输出 $/M	上下文
1	1–5	claude-opus-5.5-max	Anthropic	Proprietary	1777.4±48.9	223	$4	$20	1M
2	1–6	gpt-6-astra-max	OpenAI	Proprietary	1733.1±31.3	481	$10	$50	1M
3	1–9	claude-sonnet-5.5-xhigh	Anthropic	Proprietary	1722±44	252	$2	$10	1M
4	1–11	gpt-6.1-sol-max	OpenAI	Proprietary	1711.6±45.6	201	$2	$10	1M
5	1–9	claude-fable-5.1-max	Anthropic	Proprietary	1707.2±27.5	599	$10	$50	1M
6	2–18	gemini-4-argon-high	Google	Proprietary	1669.6±50.7	160	$2	$10	1M
7	3–16	claude-sonnet-5.5-high	Anthropic	Proprietary	1666.5±32.7	410	$2	$10	1M
8	3–13	claude-opus-5-max	Anthropic	Proprietary	1663.8±16.8	1.6k	$5	$25	1M
9	3–17	gpt-6-sol-max	OpenAI	Proprietary	1660±34.4	369	$2	$10	1M
10	6–17	claude-opus-5-high	Anthropic	Proprietary	1639.8±14.9	2.1k	$5	$25	1M
11	5–26	qwen3.8-max	Alibaba	Proprietary	1635.9±41.2	274	$2	$6	1M
12	6–18	qwen3.8-max-0902	Alibaba	Proprietary	1631.3±19.5	1.1k	$2	$6	-
13	7–26	muse-spark-1.3-max	Meta	Proprietary	1617.8±22.5	792	$1.3	$4.3	1M
14	8–25	kimi-k3-max	Moonshot	Kimi K3 license	1613.8±15.9	1.9k	$3	$15	1M
15	7–31	grok-4.7-xhigh	SpaceXAI	Proprietary	1612.1±29.9	451	$2	$6	500K
16	5–42	muse-wonder-max	Meta	Proprietary	1608.8±57.3	119	$4	$20	1M
17	10–29	claude-fable-5-high	Anthropic	Proprietary	1606±16.6	1.6k	$10	$50	1M
18	7–38	deepseek-v4.1-flash-max	DeepSeek	MIT	1601.1±36.2	280	$0.30	$1.2	1M
19	12–31	gpt-5.6-sol-xhigh (codex-harness)	OpenAI	Proprietary	1594.5±16	1.7k	$4	$20	1M
20	12–39	glm-5.3-max	Z.ai	MIT	1585.4±21.6	829	$1.4	$4.4	1M
21	12–40	muse-spark-1.3 (xHigh)	Meta	Proprietary	1583.4±23.8	681	$1.3	$4.3	1M
22	12–42	mimo-v2.6-pro	Xiaomi	MIT	1581.8±29.5	430	$0.43	$0.87	1M
23	12–41	qwen3.8-flash-next	Alibaba	qwen-community-1.0	1579.6±25.4	606	$0.16	$0.47	1M
24	12–43	mimo-v2.6-flash	Xiaomi	MIT	1578.2±30.5	431	$0.14	$0.28	1M
25	15–41	gemini-3.7-flash-high	Google	Proprietary	1574.3±19.4	1.1k	$0.75	$3.8	1M
26	13–42	grok-4.6-high	SpaceXAI	Proprietary	1573.5±21.8	803	$2	$6	500K
27	16–41	glm-5.2-max	Z.ai	MIT	1569.7±15.8	1.7k	$1.4	$4.4	1M
28	12–47	claude-haiku-5.5-high	Anthropic	Proprietary	1567.6±36.6	283	$0.10	$0.50	1M
29	15–47	hy4-preview	Tencent	Apache 2.0	1562.9±29.9	423	$0.83	$2.5	1M
30	15–48	deepseek-v4-flash-high	DeepSeek	MIT	1561.9±31.8	394	$0.30	$1.2	1M
31	18–44	gpt-6-luna-max	OpenAI	Proprietary	1561.1±15.9	1.7k	$0.10	$0.50	1M
32	18–45	qwen3.8-27b	Alibaba	Apache 2.0	1557.5±17.3	1.4k	$0.40	$3	-
33	18–47	gemini-3.8-flash-high	Google	Proprietary	1554.1±21	907	$0.75	$3.8	1M
34	18–45	claude-opus-4-8-high	Anthropic	Proprietary	1553.2±14	2.1k	$5	$25	1M
35	18–46	grok-4.5	SpaceXAI	Proprietary	1552.5±18.9	1.1k	$2	$6	500K
36	19–47	glm-5.3-flash	Z.ai	MIT	1547.6±16.6	1.4k	$0.06	$0.20	1M
37	18–56	deepseek-v4-pro-high-20260813	DeepSeek	MIT	1546.2±30.8	398	$1.3	$4.0	-
38	21–47	claude-opus-4-7-high	Anthropic	Proprietary	1544.6±13.8	2.2k	$5	$25	1M
39	16–63	muse-spark-1.2 (xHigh)	Meta	Proprietary	1543.3±45.1	166	$1.3	$4.3	-
40	20–55	muse-spark-1.1	Meta	Proprietary	1538.4±21.4	872	$1.3	$4.3	-
41	24–49	claude-opus-4-7	Anthropic	Proprietary	1538.4±14.1	2.2k	$5	$25	1M
42	27–50	claude-opus-4-6-high	Anthropic	Proprietary	1536.3±13.7	2.3k	$5	$25	1M
43	18–65	mistral-large-4	Mistral	Proprietary	1530.2±41.8	225	$0.68	$2.1	1M
44	29–57	claude-opus-4-8	Anthropic	Proprietary	1525.7±14.6	1.9k	$5	$25	1M
45	28–60	gemini-3.6-flash-high	Google	Proprietary	1525.4±20.9	931	$0.75	$3.8	1M
46	38–62	claude-opus-4-6	Anthropic	Proprietary	1515.7±13	2.6k	$5	$25	1M
47	40–63	claude-sonnet-4-6	Anthropic	Proprietary	1508.2±12.4	2.7k	$1.5	$7.5	1M
48	39–64	claude-sonnet-5-high	Anthropic	Proprietary	1507.5±16.7	1.5k	$2	$10	1M
49	37–67	hy3	Tencent	Apache 2.0	1504.5±26.5	548	$0.13	$0.53	262K
50	40–65	gpt-5.6-terra-xhigh (codex-harness)	OpenAI	Proprietary	1504±17.9	1.3k	$2	$12	1M
51	31–68	step-5-preview-high	StepFun	Proprietary	1503.2±33.1	352	$1	$2.7	1M
52	40–67	claude-opus-4-5-20251101-high-32k	Anthropic	Proprietary	1499.9±20.6	1.4k	$5	$25	200K
53	40–67	glm-5.1	Z.ai	MIT	1499.7±18.6	1.2k	$1.4	$4.4	203K
54	41–67	gpt-5.6-luna-xhigh (codex-harness)	OpenAI	Proprietary	1497.1±18.5	1.1k	$0.20	$1.2	1M
55	42–67	qwen3.7-max-20260517	Alibaba	Proprietary	1495.3±19.6	1.0k	$1.5	$4.4	1M
56	43–67	gpt-5.5-xhigh (codex-harness)	OpenAI	Proprietary	1494.1±14.8	1.9k	$5	$30	-
57	43–67	seed-2.1-pro-preview	Bytedance	Proprietary	1494±16.8	1.4k	$0.88	$4.4	262K
58	43–68	gemini-3.5-flash-medium	Google	Proprietary	1489.3±17.8	1.2k	$0.75	$4.5	1M
59	44–68	kimi-k2.6	Moonshot	Modified MIT	1487±17.5	1.3k	$0.95	$4	262K
60	32–84	gpt-5.4-medium (codex-harness)	OpenAI	Proprietary	1483.8±49.8	156	$2.5	$15	1M
61	45–73	gemini-3.5-flash-high	Google	Proprietary	1478.8±19.7	1.0k	$0.75	$4.5	1M
62	48–73	gpt-5.5-high (codex-harness)	OpenAI	Proprietary	1475.8±13.5	2.3k	$5	$30	-
63	40–89	gemini-3.5-flash	Google	Proprietary	1473.2±49.2	162	$0.75	$4.5	1M
64	50–73	mimo-v2.5-pro	Xiaomi	MIT	1470.6±13.1	2.4k	$0.43	$0.87	1M
65	50–74	minimax-m3	MiniMax	MiniMax Community License	1470.1±14.4	2.0k	$0.30	$1.2	524K
66	47–88	qwen3.6-max-preview	Alibaba	Proprietary	1458.9±33.1	365	$1.3	$7.6	262K
67	60–83	qwen3.6-plus	Alibaba	Proprietary	1454.4±13.5	2.2k	$0.33	$1.9	1M
68	60–84	deepseek-v4-pro-high-preview	DeepSeek	MIT	1453.7±15.8	1.6k	$1.3	$4.0	1M
69	57–89	kimi-k2.7-code	Moonshot	Modified MIT	1448.9±25.7	636	$0.67	$3.4	262K
70	60–86	claude-opus-4-5-20251101	Anthropic	Proprietary	1446.7±17.6	1.7k	$5	$25	200K
71	44–103	gpt-5.4-high (codex-harness)	OpenAI	Proprietary	1446.4±57.3	138	$2.5	$15	1M
72	64–89	gpt-5.5 (codex-harness)	OpenAI	Proprietary	1436.5±14.7	1.9k	$5	$30	-
73	63–92	minimax-m2.5	MiniMax	Modified MIT	1435.9±20.4	1.0k	$0.27	$1.1	205K
74	64–89	gemini-3.1-pro-preview	Google	Proprietary	1435.6±11.9	3.0k	$1	$6	1M
75	64–91	mimo-v2.5	Xiaomi	MIT	1434.2±14.8	1.8k	$0.14	$0.28	1M
76	64–92	deepseek-v4-pro	DeepSeek	MIT	1431.7±14.6	1.8k	$1.3	$4.0	1M
77	60–103	gpt-5.3-codex (codex-harness)	OpenAI	Proprietary	1431.4±37.9	338	$1.8	$14	400K
78	64–92	kimi-k2.5-thinking	Moonshot	Modified MIT	1429.3±13.1	2.4k	$0.60	$3	-
79	64–95	mimo-v2-pro	Xiaomi	Proprietary	1427.8±21.9	869	$1.1	$3.3	1M
80	64–99	glm-5	Z.ai	MIT	1421.7±21.1	962	$1	$3.2	203K
81	64–104	glm-4.7	Z.ai	MIT	1414.7±31.1	514	$0.60	$2.2	205K
82	64–104	gpt-5.4	OpenAI	Proprietary	1411.9±30.4	440	$2.5	$15	1M
83	67–103	gemini-3-flash	Google	Proprietary	1410.3±19.7	1.3k	$0.50	$3	1M
84	68–103	deepseek-v4-flash-high-preview	DeepSeek	MIT	1409.5±17.5	1.3k	$0.30	$1.2	1M
85	73–104	gemini-3-pro	Google	Proprietary	1401.4±18.3	1.7k	$2	$12	1M
86	67–110	gpt-5.1-medium	OpenAI	Proprietary	1399.4±32.3	522	$0.63	$5	400K
87	77–104	minimax-m2.7	MiniMax	Modified MIT	1395.9±15.2	1.8k	$0.30	$1.2	205K
88	60–120	gpt-5.2	OpenAI	Proprietary	1395.8±71.3	111	$0.88	$7	400K
89	65–116	gpt-5-medium	OpenAI	Proprietary	1394.1±44.8	297	$1.3	$10	400K
90	69–112	kimi-k2.5-instant	Moonshot	Modified MIT	1393.7±31.7	428	$0.50	$2.5	262K
91	78–104	qwen3.5-397b-a17b	Alibaba	Apache 2.0	1391.4±12.6	2.6k	$0.50	$3.6	262K
92	77–110	minimax-m2.1-preview	MiniMax	MIT	1390.8±23.4	953	$0.30	$1.2	205K
93	68–116	glm-5v-turbo	Z.ai	Proprietary	1389±37.1	299	$1.2	$4	203K
94	78–108	gpt-5.4-mini-high	OpenAI	Proprietary	1388.7±16.6	1.5k	$0.75	$4.5	400K
95	74–114	gpt-5.3-codex (codex-harness)	OpenAI	Proprietary	1387.7±31.4	441	$1.8	$14	400K
96	79–110	claude-sonnet-4-5-20250929	Anthropic	Proprietary	1383.1±16.7	1.9k	$3	$15	1M
97	79–111	inkling	Thinky	Apache 2.0	1381.3±17.8	1.3k	$1	$4.0	524K
98	79–112	claude-sonnet-4-5-20250929-high-32k	Anthropic	Proprietary	1379.7±18.3	1.6k	$3	$15	200K
99	78–116	claude-opus-4-1-20250805	Anthropic	Proprietary	1375.5±27.9	739	$7.5	$37.5	200K
100	78–118	Inkling Small	Thinky	Apache 2.0	1369±31.9	397	$0.45	$1.2	524K
101	73–123	muse-glimmer	Meta	Apache-2.0	1366.8±54.6	131	-	-	-
102	88–116	gemini-3-flash (thinking-minimal)	Google	Proprietary	1365.8±12.5	2.8k	$0.50	$3	1M
103	83–118	deepseek-v3.2-thinking	DeepSeek	MIT	1364.4±22.8	886	$0.26	$0.80	164K
104	77–122	hunyuan-hy3-preview	Tencent	tencent-hunyuan-community	1362.6±48.5	181	-	-	-
105	79–120	solar-pro4	Upstage	Proprietary	1359.7±34.3	371	$0.09	$0.36	524K
106	88–118	qwen3.5-122b-a10b	Alibaba	Apache 2.0	1356.5±20	1.0k	$0.26	$2.1	262K
107	88–118	qwen3.5-27b	Alibaba	Apache 2.0	1355.9±21.1	964	$0.26	$2.6	262K
108	89–118	grok-4.20-beta-0309-reasoning	SpaceXAI	Proprietary	1355.5±14	2.1k	$1.3	$2.5	1M
109	88–120	laguna-m.1	Poolside	Apache 2.0	1349.7±25.9	686	$0.20	$0.40	262K
110	89–122	mimo-v2-flash (non-thinking)	Xiaomi	MIT	1343.3±24.8	764	$0.10	$0.30	262K
111	95–120	grok-4.3	SpaceXAI	Proprietary	1342.4±15.5	1.8k	$1.3	$2.5	1M
112	93–122	gpt-5.2-codex	OpenAI	Proprietary	1339.9±22.9	895	$1.8	$14	400K
113	92–123	glm-4.6	Z.ai	MIT	1338.1±27.9	706	$0.50	$2	205K
114	96–122	gpt-5.1	OpenAI	Proprietary	1335±20.2	1.3k	$0.63	$5	400K
115	96–122	gemma-4-31b	Google	Apache 2.0	1334.2±19.3	1.3k	$0.99	$1.5	131K
116	100–123	kimi-k2-thinking-turbo	Moonshot	Modified MIT	1324.1±18.3	1.5k	$0.57	$2.3	131K
117	105–123	claude-haiku-4-5-20251001	Anthropic	Proprietary	1320.6±11.9	3.3k	$1	$5	200K
118	100–128	gpt-5.1-codex	OpenAI	Proprietary	1312.9±32.1	544	$1.3	$10	400K
119	105–125	deepseek-v3.2	DeepSeek	MIT	1311.8±19.9	1.2k	$0.26	$0.80	164K
120	95–133	gemma-4-26b-a4b	Google	Apache 2.0	1305.2±53.3	156	-	-	-
121	109–130	minimax-m2	MiniMax	Apache 2.0	1294.2±26.9	722	$0.30	$1.2	205K
122	114–133	laguna-xs.2	Poolside	Apache 2.0	1284.5±28.5	553	$0.10	$0.20	262K
123	109–135	mistral-medium-3.5	Mistral	Modified MIT	1282.3±39.3	295	$0.75	$3.8	262K
124	119–134	qwen3-coder-480b-a35b-instruct	Alibaba	Apache 2.0	1263.4±19	1.5k	$0.65	$3.3	1M
125	119–136	grok-4-1-fast-reasoning	SpaceXAI	Proprietary	1254.5±27	704	$0.20	$0.50	128K
126	118–137	mimo-v2-flash (thinking)	Xiaomi	MIT	1245.4±47.3	207	$0.10	$0.30	262K
127	120–137	deepseek-v3.2-exp	DeepSeek	MIT	1241.7±36.4	419	$0.27	$0.41	164K
128	121–136	gemini-3.1-flash-lite-preview	Google	Proprietary	1240.5±16.5	1.9k	$0.25	$1.5	1M
129	119–138	gemini-2.5-pro	Google	Proprietary	1237.2±44.4	282	$1.3	$10	1M
130	118–138	grok-4.1-thinking	SpaceXAI	Proprietary	1224.9±71.9	113	-	-	-
131	121–138	qwen3.5-35b-a3b	Alibaba	Apache 2.0	1219.7±47.5	202	$0.15	$1	262K
132	120–138	trinity-large-thinking	-	Apache 2.0	1204.3±64	130	$0.25	$0.80	262K
133	121–138	gpt-5.1-codex-mini	OpenAI	Proprietary	1199.3±66.8	123	$0.25	$2	400K
134	123–138	qwen3.5-flash	Alibaba	Proprietary	1192.1±56.1	174	-	-	-
135	124–138	KAT-Coder-Pro-V1	-	Proprietary	1172.7±71.7	142	$0.21	$0.83	256K
136	125–138	grok-code-fast-1	SpaceXAI	Proprietary	1164.2±74.9	92	$0.20	$1.5	-
137	127–138	devstral-2	Mistral	Modified MIT	1163.1±58.8	145	-	-	-
138	129–138	granite-4.1-8b	IBM	Apache 2.0	1137.3±55.6	212	$0.05	$0.10	131K

Elo 分真人在不知道模型名字的情况下两两投票，胜负累积出的等级分。绝对值没有意义，差值才有：分差 100 分 ≈ 强的那个有 64% 胜率，只差几分基本是平手。
± 后面的数是 95% 置信区间的半宽，投票越多越小。两个模型的区间重叠，说明它们的名次差别在统计上不成立。
区间把上面这个置信区间换算成名次：1–5 意思是这个模型的真实名次可能在第 1 到第 5 之间。

数据更新: 2026-10-14数据来源: Arena AI ↗