Skip to content

Best AI for writing: the answers people prefer

Blind votes on which of two answers people prefer (Text Arena Elo), plus language comprehension (LiveBench).

DataText Arena8 Oct 2026outdatedfetched

Text Arena leaderboard, 414 models
#ModelScore
1
1525Elo ±9
2
1507Elo ±8
3
1504Elo ±3
4
1504Elo ±4
5
1501Elo ±4
6
1501Elo ±6
7
1498Elo ±3
8
1497Elo ±5
9
1494Elo ±6
10
1494Elo ±4
11
1492Elo ±8
12
1491Elo ±4
13
1490Elo ±4
14
1489Elo ±5
15
1489Elo ±6
16
Kimi K3 MaxMoonshot AI
1488Elo ±5
17
1487Elo ±3
18
1486Elo ±5
19
1485Elo ±4
20
1485Elo ±4
21
1484Elo ±9
22
1483Elo ±5
23
1482Elo ±4
24
1482Elo ±4
25
1481Elo ±4
26
1480Elo ±9
27
1478Elo ±5
28
1478Elo ±4
29
GPT-5.5OpenAI
1477Elo ±4
30
1476Elo ±4
31
1476Elo ±4
32
1476Elo ±9
33
1475Elo ±4
34
1475Elo ±4
35
1475Elo ±7
36
1475Elo ±10
37
1475Elo ±5
38
1475Elo ±5
39
1475Elo ±7
40
1474Elo ±4
41
1474Elo ±4
42
1473Elo ±5
43
1473Elo ±4
44
1472Elo ±4
45
1472Elo ±4
46
1470Elo ±4
47
1470Elo ±3
48
1468Elo ±4
49
1467Elo ±5
50
1466Elo ±4
51
Grok 4.5SpaceXAI
1466Elo ±4
52
1465Elo ±3
53
GPT-5.4OpenAI
1465Elo ±4
54
1465Elo ±5
55
1465Elo ±7
56
1464Elo ±4
57
1461Elo ±4
58
Kimi K2.6Moonshot AI
1461Elo ±4
59
1460Elo ±8
60
Grok 4.1SpaceXAI
1459Elo ±3
61
1458Elo ±3
62
1458Elo ±4
63
GLM-5Z.AI
1458Elo ±4
64
1456Elo ±3
65
1456Elo ±3
66
1456Elo ±4
67
1456Elo ±7
68
1456Elo ±4
69
1455Elo ±3
70
1455Elo ±4
71
1455Elo ±4
72
1454Elo ±5
73
Hy3Tencent
1454Elo ±6
74
1452Elo ±4
75
1452Elo ±7
76
1451Elo ±7
77
1450Elo ±4
78
1450Elo ±3
79
1450Elo ±3
80
1449Elo ±7
81
1448Elo ±3
82
1448Elo ±5
83
1447Elo ±9
84
1447Elo ±4
85
1446Elo ±4
86
1445Elo ±2
87
1445Elo ±6
88
1443Elo ±4
89
1443Elo ±3
90
1443Elo ±7
91
1443Elo ±5
92
1442Elo ±3
93
Grok 4.3SpaceXAI
1442Elo ±4
94
1441Elo ±6
95
InklingThinking Machines
1441Elo ±4
96
MiniMax M3MiniMax
1440Elo ±4
97
1439Elo ±4
98
GPT-5.1OpenAI
1439Elo ±4
99
1438Elo ±5
100
1438Elo ±7
101
1437Elo ±4
102
1436Elo ±5
103
1436Elo ±4
104
GPT-5.2OpenAI
1435Elo ±3
105
1435Elo ±5
106
1434Elo ±5
107
MiMo V2.5Xiaomi
1434Elo ±4
108
1433Elo ±6
109
1433Elo ±4
110
1432Elo ±4
111
1431Elo ±6
112
1430Elo ±3
113
1430Elo ±7
114
1430Elo ±3
115
1429Elo ±9
116
1426Elo ±10
117
1426Elo ±6
118
1426Elo ±4
119
1426Elo ±4
120
1426Elo ±7
121
1425Elo ±7
122
1425Elo ±3
123
1425Elo ±10
124
1424Elo ±4
125
1423Elo ±6
126
1423Elo ±4
127
1422Elo ±3
128
1421Elo ±6
129
1421Elo ±6
130
1419Elo ±8
131
1419Elo ±9
132
1418Elo ±7
133
1418Elo ±5
134
1417Elo ±6
135
1416Elo ±4
136
1416Elo ±10
137
1416Elo ±7
138
1415Elo ±3
139
1415Elo ±10
140
1415Elo ±4
141
1414Elo ±4
142
1414Elo ±2
143
1414Elo ±3
144
1413Elo ±7
145
1413Elo ±10
146
1411Elo ±4
147
Grok 4 0709SpaceXAI
1411Elo ±4
148
1411Elo ±5
149
1409Elo ±2
150
1409Elo ±7
151
1409Elo ±4
152
1408Elo ±3
153
Inkling SmallThinking Machines
1405Elo ±5
154
1405Elo ±5
155
1403Elo ±4
156
1402Elo ±4
157
1402Elo ±5
158
1402Elo ±4
159
1401Elo ±4
160
1401Elo ±6
161
1400Elo ±7
162
1399Elo ±5
163
DeepSeek R1DeepSeek
1398Elo ±5
164
1396Elo ±4
165
1395Elo ±4
166
1395Elo ±7
167
1394Elo ±4
168
1394Elo ±12
169
1393Elo ±4
170
1393Elo ±9
171
1392Elo ±3
172
1391Elo ±4
173
1391Elo ±4
174
1391Elo ±4
175
1390Elo ±5
176
1389Elo ±5
177
1388Elo ±4
178
1388Elo ±5
179
1387Elo ±5
180
1387Elo ±9
181
1386Elo ±6
182
Solar Pro4Upstage
1386Elo ±6
183
1385Elo ±5
184
1383Elo ±6
185
1383Elo ±4
186
1382Elo ±5
187
1379Elo ±3
188
1379Elo ±11
189
1379Elo ±4
190
1375Elo ±5
191
1375Elo ±5
192
1374Elo ±3
193
1374Elo ±4
194
1373Elo ±4
195
1372Elo ±4
196
1369Elo ±6
197
1367Elo ±5
198
1365Elo ±4
199
1365Elo ±6
200
1364Elo ±4
201
1364Elo ±5
202
MiniMax M1MiniMax
1364Elo ±4
203
1364Elo ±5
204
1360Elo ±7
205
1360Elo ±4
206
DeepSeek V3DeepSeek
1358Elo ±5
207
1358Elo ±5
208
1356Elo ±5
209
1356Elo ±8
210
1354Elo ±3
211
1354Elo ±4
212
1352Elo ±8
213
1352Elo ±4
214
1351Elo ±3
215
1349Elo ±6
216
Step 3Stepfun
1349Elo ±8
217
1349Elo ±12
218
o3 MiniOpenAI
1348Elo ±4
219
1348Elo ±6
220
1348Elo ±12
221
Qwen 3 32BAlibaba
1347Elo ±9
222
1346Elo ±11
223
1346Elo ±3
224
1346Elo ±8
225
Mercury 2Inception AI
1344Elo ±11
226
1344Elo ±7
227
MiniMax M2MiniMax
1343Elo ±8
228
1343Elo ±3
229
1343Elo ±10
230
1343Elo ±8
231
1342Elo ±10
232
1341Elo ±12
233
1340Elo ±10
234
1337Elo ±7
235
o1 MiniOpenAI
1337Elo ±4
236
1336Elo ±5
237
1336Elo ±4
238
Qwq 32BAlibaba
1336Elo ±4
239
1336Elo ±4
240
1335Elo ±4
241
1335Elo ±6
242
1334Elo ±4
243
1334Elo ±9
244
1329Elo ±6
245
1328Elo ±5
246
1328Elo ±12
247
1327Elo ±4
248
1327Elo ±5
249
1326Elo ±10
250
1325Elo ±3
251
1324Elo ±4
252
1324Elo ±4
253
1323Elo ±8
254
1323Elo ±7
255
1322Elo ±8
256
1322Elo ±3
257
1322Elo ±21
258
1322Elo ±5
259
1320Elo ±7
260
1319Elo ±5
261
1318Elo ±6
262
1318Elo ±4
263
1318Elo ±4
264
1318Elo ±5
265
1317Elo ±6
266
1315Elo ±6
267
1314Elo ±4
268
1314Elo ±5
269
1314Elo ±5
270
1313Elo ±4
271
1313Elo ±4
272
1311Elo ±10
273
1309Elo ±4
274
1308Elo ±4
275
1307Elo ±5
276
1307Elo ±6
277
1306Elo ±8
278
MercuryInception AI
1306Elo ±14
279
1306Elo ±4
280
1305Elo ±6
281
1304Elo ±10
282
1303Elo ±9
283
1303Elo ±4
284
1303Elo ±5
285
1299Elo ±8
286
1294Elo ±9
287
1293Elo ±4
288
1290Elo ±11
289
1290Elo ±5
290
1290Elo ±12
291
1289Elo ±3
292
1289Elo ±7
293
1288Elo ±7
294
1288Elo ±5
295
1287Elo ±8
296
1287Elo ±7
297
1287Elo ±10
298
1287Elo ±5
299
1286Elo ±10
300
1281Elo ±4
301
1280Elo ±7
302
1277Elo ±5
303
1277Elo ±4
304
1276Elo ±7
305
1276Elo ±4
306
1274Elo ±6
307
1273Elo ±7
308
1272Elo ±7
309
1271Elo ±8
310
1267Elo ±5
311
1267Elo ±4
312
1265Elo ±6
313
1262Elo ±5
314
1262Elo ±4
315
1262Elo ±4
316
1260Elo ±5
317
1259Elo ±4
318
Phi 4Microsoft
1256Elo ±5
319
1252Elo ±11
320
1250Elo ±7
321
1243Elo ±5
322
1241Elo ±5
323
1240Elo ±7
324
1238Elo ±9
325
1237Elo ±7
326
1234Elo ±6
327
1234Elo ±7
328
1234Elo ±5
329
1233Elo ±12
330
1230Elo ±5
331
Command RCohere
1227Elo ±5
332
1227Elo ±6
333
1226Elo ±5
334
1224Elo ±4
335
1224Elo ±12
336
1223Elo ±7
337
1223Elo ±5
338
1221Elo ±11
339
1213Elo ±11
340
1213Elo ±5
341
1211Elo ±4
342
1209Elo ±11
343
1204Elo ±9
344
1204Elo ±6
345
1200Elo ±4
346
1198Elo ±5
347
1197Elo ±4
348
1195Elo ±6
349
1191Elo ±7
350
1191Elo ±7
351
1185Elo ±11
352
Wizardlm 70BMicrosoft
1185Elo ±9
353
1184Elo ±7
354
1183Elo ±9
355
1183Elo ±10
356
1183Elo ±8
357
1183Elo ±6
358
1180Elo ±6
359
1179Elo ±11
360
1178Elo ±10
361
1176Elo ±10
362
Vicuna 33BUnknown
1173Elo ±6
363
1171Elo ±6
364
1171Elo ±7
365
1171Elo ±5
366
1167Elo ±8
367
1167Elo ±8
368
1164Elo ±12
369
1157Elo ±8
370
1155Elo ±13
371
1154Elo ±11
372
1152Elo ±13
373
1152Elo ±15
374
1151Elo ±12
375
Wizardlm 13BMicrosoft
1150Elo ±9
376
1149Elo ±7
377
1148Elo ±17
378
1144Elo ±10
379
1143Elo ±6
380
Vicuna 13BUnknown
1142Elo ±7
381
1141Elo ±7
382
1139Elo ±11
383
Palm 2Google
1139Elo ±9
384
1138Elo ±9
385
1137Elo ±9
386
1131Elo ±9
387
1130Elo ±7
388
1128Elo ±6
389
1127Elo ±12
390
1127Elo ±16
391
1122Elo ±11
392
1119Elo ±18
393
1117Elo ±8
394
Vicuna 7BUnknown
1115Elo ±9
395
1115Elo ±14
396
1111Elo ±8
397
1110Elo ±9
398
1108Elo ±7
399
1094Elo ±11
400
1091Elo ±9
401
1074Elo ±11
402
Koala 13BUnknown
1071Elo ±10
403
Alpaca 13BUnknown
1070Elo ±11
404
1068Elo ±15
405
1063Elo ±12
406
1057Elo ±12
407
1042Elo ±11
408
1025Elo ±14
409
1023Elo ±11
410
996Elo ±12
411
993Elo ±12
412
982Elo ±13
413
975Elo ±16
414
953Elo ±13

Need a model picked for your use case?

Benchmarks rank models on other people's tasks. For a business workflow I test the shortlist on your own data, costs and privacy requirements, and recommend one.

AI consulting