OnlySOTA

Arena — Vision

按 Arena 发布的原样呈现,指标和模型都以它为准。

发布方
Arena
许可协议
未声明
更新频率
持续更新
数据截至
2026-10-09
本站抓取时间
2026-10-10
条目数
165 · 11 条未识别
按 Overall (style control) 排名前 12 的模型
  1. Claude Fable 5 1308
  2. Qwen 3.8 Max 1301
  3. Claude Opus 4.6 1299
  4. Claude Opus 4.7 1298
  5. Gemini 3.7 Flash 1297
  6. GPT 6.1 Sol 1294
  7. Muse Spark 1294
  8. Muse Spark 1.2 1293
  9. Claude Opus 5.5 1293
  10. Muse Spark 1.3 1290
  11. Gemini 3 Pro 1289
  12. Gemini 3.8 Flash 1289

Elo 分没有零点,所以每行是从 1287 起的刻度上的一个点:看点与点之间的距离,不看离左边多远。

165 个有分数

#模型
1Claude Fable 5highAnthropic13081301–1315
2Qwen 3.8 MaxAlibaba13011293–1308
3Claude Opus 4.6highAnthropic12991293–1306
4Claude Opus 4.7Anthropic12981292–1305
5Claude Opus 4.7highAnthropic12981291–1304
6Gemini 3.7 FlashhighGoogle12971287–1307
7Claude Opus 4.6Anthropic12941288–1301
8GPT 6.1 SolmaxOpenAI12941280–1308
9Muse SparkMeta12941285–1303
10Muse Spark 1.2xhighMeta12931279–1307
11Claude Opus 5.5highAnthropic12931280–1305
12Muse Spark 1.3maxMeta12901280–1300
13Gemini 3 ProGoogle12891282–1297
14Gemini 3.8 FlashhighGoogle12891279–1300
15Claude Opus 5highAnthropic12871281–1294
16GPT 5.6 SolxhighOpenAI12871280–1294
17Claude Fable 5.1maxAnthropic12861277–1296
18Muse Spark 1.3xhighMeta12861276–1296
19GPT 6 AstramaxOpenAI12861275–1297
20GPT 5.5OpenAI12861280–1292
21Claude Opus 4.8highAnthropic12861279–1292
22Gemini 3.5 FlashmediumGoogle12851278–1292
23GPT 5.4highOpenAI12851279–1291
24Gemini 3.5 FlashhighGoogle12841277–1291
25GPT 5.5highOpenAI12811275–1287
26Gemini 3.6 FlashhighGoogle12811273–1289
27Grok 4.5SpaceXAI12801273–1287
28Muse Spark 1.1Meta12801273–1287
29Gemini 3.1 PropreviewGoogle12801275–1285
30Claude Opus 4.8Anthropic12791272–1285
31GPT 5.4OpenAI12791272–1285
32GPT 5.220260210 · chatOpenAI12781271–1284
33GPT 5.5 InstantOpenAI12761267–1284
34Claude Sonnet 4.6Anthropic12751269–1281
35GLM 5.3 FlashZ AI12731266–1280
36Gemini 3 FlashGoogle12731268–1278
37GPT 6 SolmaxOpenAI12701258–1283
38GPT 5.6 TerraxhighOpenAI12681261–1275
39Claude Sonnet 5.5xhighAnthropic12661252–1280
40Gemini 3.5 Flash LiteGoogle12661258–1274
41Kimi K2.6Kimi12651259–1272
42Qwen 3.7 PlusAlibaba12641257–1271
43Claude Sonnet 5highAnthropic12631256–1270
44Gemma 431BGoogle12621256–1268
45Grok 4.6highSpaceXAI12621253–1271
46DeepSeek V4.1 FlashmaxDeepSeek12601249–1271
47Gemini 3 Flashon · minimalGoogle12601255–1265
48GPT 5.6 LunaxhighOpenAI12581250–1265
49dola seed 2.0 proByteDance12561249–1263
50Grok 4.20beta · 0309 · onSpaceXAI12561249–1262
51Kimi K2.5onKimi12521246–1258
52GPT 5.1highOpenAI12511244–1259
53GPT 5.4 minihighOpenAI12511245–1257
54Grok 4.20beta · 0309 · multi-agentSpaceXAI12501244–1257
55Step 5previewStepFun12501232–1268
56MiMo V2.6 FlashXiaomi12481236–1260
57Gemini 2.5 ProGoogle12481243–1252
58Grok 4.7xhighSpaceXAI12481236–1259
59MiMo V2.6 ProXiaomi12461231–1261
60Qwen 3.5397B A17BAlibaba12461241–1251
61Qwen 3.827BAlibaba12441236–1253
62GPT 5.2highOpenAI12431237–1249
63Grok 4.3SpaceXAI12431237–1248
64Gemma 426B A4BGoogle12421235–1248
65GPT 6 LunamaxOpenAI12411229–1253
66GPT 5.1OpenAI12401232–1247
67GPT 4o (ChatGPT)20250326OpenAI12391233–1244
68Kimi K2.5Kimi12361226–1247
69MiniMax M3MiniMax12361229–1242
70Gemini 3.1 Flash LitepreviewGoogle12351229–1240
71MiMo V2.5Xiaomi12341228–1239
72GLM 5V TurboZ AI12311225–1236
73Gemini 2.5 Flashpreview · 09-2025Google12291219–1239
74GPT 5.2OpenAI12291223–1235
75Qwen 3.5122B A10BAlibaba12261219–1233
76GPT 5OpenAI12261219–1233
77GPT 4.5preview · 2025-02-27OpenAI12251213–1237
78ERNIE 5.0preview · 1220Baidu12191208–1231
79Qwen 3.527BAlibaba12171211–1222
80MiMo V2 OmniXiaomi12161209–1224
81o32025-04-16OpenAI12161209–1222
82Gemini 2.5 FlashGoogle12151211–1220
83Qwen 3 VL235B A22BAlibaba12151208–1222
84GPT 4.12025-04-14OpenAI12141207–1221
85GPT 5highOpenAI12101203–1217
86Inkling SmallThinking Machines12081199–1216
87Claude Opus 420250514 · on · 16000Anthropic12081191–1224
88Claude Sonnet 420250514 · on · 32000Anthropic12061189–1223
89GPT 4.1 mini2025-04-14OpenAI12011193–1208
90o4 mini2025-04-16OpenAI12011194–1208
91Mistral Medium 3.5Mistral12001191–1210
92Mistral Large 3Mistral12001192–1208
93GPT 5.4 nanohighOpenAI11991193–1206
94Grok 4.1 FastonSpaceXAI11961189–1203
95Claude Sonnet 3.720250219 · on · 32000Anthropic11931177–1208
96o12024-12-17OpenAI11931182–1203
97Qwen 3 VL235B A22B · onAlibaba11891176–1201
98Gemini 2.5 Flash Litepreview · 06-17 · onGoogle11881181–1196
99Claude Opus 420250514Anthropic11881176–1201
100Claude Sonnet 420250514Anthropic11881174–1202

每一列测的是什么

Overall (style control)
越高越好 · 多模态