OnlySOTA

Arena — Text

照 Arena 发布的原样呈现:它的指标,它的模型。点模型名进本站详情,点箭头回到原榜。

发布方
Arena
许可协议
未声明
更新频率
持续更新
数据截至
2026-10-02
本站抓取时间
2026-10-02
条目数
413 · 44 条未识别
按 Overall (style control) 排名前 12 的模型
  1. Gemini 4 Argon 1525
  2. Claude Opus 4.6 1505
  3. Claude Fable 5 1504
  4. Claude Opus 5.5 1504
  5. Claude Opus 4.7 1501
  6. Claude Fable 5.1 1501
  7. Gemini 3.8 Flash 1495
  8. Muse Spark 1.3 1494
  9. Muse Spark 1.2 1494
  10. Muse Spark 1.1 1491
  11. Claude Opus 5 1490
  12. Muse Spark 1489

Elo 分没有真正的零点,所以这里不画从 0 开始的条形,而是在从 1485 起的刻度上画点。请看点与点之间的距离,而不是离左边多远。

413 个有分数

#模型
1Gemini 4 ArgonhighGoogle15251516–1534
2Claude Opus 4.6highAnthropic15051501–1508
3Claude Fable 5highAnthropic15041500–1509
4Claude Opus 5.5highAnthropic15041495–1513
5Claude Opus 4.7highAnthropic15011498–1505
6Claude Fable 5.1maxAnthropic15011495–1508
7Claude Opus 4.6Anthropic14971494–1501
8Gemini 3.8 FlashhighGoogle14951490–1500
9Muse Spark 1.3maxMeta14941488–1500
10Claude Opus 4.7Anthropic14941490–1498
11Muse Spark 1.2xhighMeta14941484–1503
12Muse Spark 1.1Meta14911487–1496
13Claude Opus 5highAnthropic14901486–1494
14Claude Opus 5maxAnthropic14891485–1494
15Muse SparkMeta14891484–1495
16Kimi K3maxKimi14881484–1493
17Gemini 3.7 FlashhighGoogle14881483–1493
18Gemini 3.1 PropreviewGoogle14871484–1490
19Gemini 3 ProGoogle14851482–1489
20GPT 5.6 SolxhighOpenAI14841479–1488
21GPT 6.1 SolmaxOpenAI14831473–1494
22Gemini 3.6 FlashhighGoogle14831479–1487
23Qwen 3.8 MaxAlibaba14821476–1487
24Claude Opus 4.8highAnthropic14821478–1485
25GPT 5.5highOpenAI14811478–1485
26MiMo V2.6 ProXiaomi14801471–1489
27GLM 5.3maxZ AI14781473–1484
28Gemini 3.5 FlashhighGoogle14771473–1481
29GPT 6 AstramaxOpenAI14771470–1484
30GPT 5.5OpenAI14771473–1480
31Gemini 3.5 FlashmediumGoogle14761472–1480
32GLM 5.2maxZ AI14761471–1480
33GPT 5.220260210 · chatOpenAI14761472–1480
34GPT 5.4highOpenAI14751471–1479
35Qwen 3.7 MaxpreviewAlibaba14751465–1485
36Grok 4.20beta1SpaceXAI14751470–1479
37Claude Opus 4.8Anthropic14751471–1478
38DeepSeek V4.1 FlashmaxDeepSeek14741468–1481
39Claude Opus 4.520251101 · high · 32000Anthropic14741470–1477
40GPT 5.5 InstantOpenAI14731468–1478
41GLM 5.3 FlashZ AI14731468–1478
42Gemini 3 FlashGoogle14731468–1477
43Claude Sonnet 4.6Anthropic14721469–1476
44Grok 4.20beta · 0309 · onSpaceXAI14721468–1475
45Claude Sonnet 5.5xhighAnthropic14711461–1481
46Grok 4.20beta · 0309 · multi-agentSpaceXAI14711467–1474
47Claude Opus 4.520251101Anthropic14701467–1473
48ERNIE 5.1Baidu14681463–1472
49MiMo V2.5 ProXiaomi14681464–1471
50Grok 4.5SpaceXAI14661461–1470
51GPT 5.6 TerraxhighOpenAI14651461–1470
52Grok 4.1onSpaceXAI14651462–1468
53GPT 5.4OpenAI14651461–1468
54Qwen 3.5 MaxpreviewAlibaba14651460–1470
55GLM 5.1Z AI14651461–1468
56DeepSeek V4 Pro20260813 · highDeepSeek14641458–1471
57Claude Sonnet 5highAnthropic14621458–1466
58Kimi K2.6Kimi14611457–1465
59Qwen 3.6 MaxpreviewAlibaba14601452–1469
60Grok 4.1SpaceXAI14591456–1462
61Gemini 3 Flashon · minimalGoogle14581455–1461
62DeepSeek V4 ProDeepSeek14581454–1462
63GLM 5Z AI14581453–1462
64GPT 6 SolmaxOpenAI14571450–1464
65Claude Sonnet 4.520250929 · high · 32000Anthropic14571454–1459
66dola seed 2.0 proByteDance14561453–1460
67Hunyuan Hy3Tencent14561449–1463
68Qwen 3.7 PlusAlibaba14561451–1460
69GPT 5.1highOpenAI14561452–1459
70Claude Sonnet 4.520250929Anthropic14551452–1458
71Gemini 3.5 Flash LiteGoogle14551450–1459
72DeepSeek V4 Propreview · highDeepSeek14551451–1459
73Grok 4.6highSpaceXAI14541449–1459
74Step 5previewStepFun14541443–1465
75GPT 5.6 LunaxhighOpenAI14531449–1457
76Gemma 431BGoogle14531445–1460
77MiMo V2.6 FlashXiaomi14521444–1460
78Kimi K2.5onKimi14501447–1453
79GPT 5.3chatOpenAI14501446–1454
80Claude Opus 4.120250805 · on · 16000Anthropic14501446–1453
81ERNIE 5.0preview · 1203Baidu14491442–1455
82Claude Opus 4.120250805Anthropic14481445–1451
83MiMo V2 ProXiaomi14481443–1453
84GPT 5.4 minihighOpenAI14471443–1451
85ERNIE 5.00110Baidu14461442–1450
86Gemini 2.5 ProGoogle14461443–1448
87GPT 4.5preview · 2025-02-27OpenAI14451439–1450
88Qwen 3.6 PlusAlibaba14431439–1448
89GPT 4o (ChatGPT)20250326OpenAI14431440–1446
90GPT 6 LunamaxOpenAI14431436–1449
91Grok 4.7xhighSpaceXAI14421435–1450
92Qwen 3.5397B A17BAlibaba14421439–1445
93GLM 4.7Z AI14411435–1448
94Grok 4.3SpaceXAI14411438–1445
95InklingThinking Machines14411437–1446
96MiniMax M3MiniMax14401436–1444
97GPT 5.1OpenAI14391435–1442
98DeepSeek V4 Flashpreview · highDeepSeek14381434–1443
99Qwen 3.827BAlibaba14381432–1443
100Gemma 426B A4BGoogle14371430–1445

每一列测的是什么

Overall (style control)
越高越好 · 综合偏好