Noometry

Reasoning benchmark

LMArena Hard Prompts leaderboard

As of October 2026, Gemini 4 Argon has the highest published LMArena Hard Prompts score on Noometry at 1548, out of 297 models with results.

Last verified

About LMArena Hard Prompts

LMArena ratings restricted to prompts classified as hard.

Category
Reasoning
Introduced
2024
Format
Pairwise human votes
Unit
Arena rating (Bradley–Terry)
Official site
lmarena.ai

Top 15 models

  1. Gemini 4 Argon 1548
  2. Claude Opus 5.5 1535
  3. Claude Opus 4.6 1527
  4. Claude Opus 5 1526
  5. Claude Fable 5.1 1526
  6. MiMo-V2.6-Pro 1512
  7. Gemini 3.8 Flash 1508
  8. Claude Fable 5 1508
  9. Claude Opus 4.7 1506
  10. Muse Spark 1.3 1503
  11. Qwen3.8 Max 1496
  12. Kimi K3 1496
  13. Claude Sonnet 5.5 1495
  14. Gemini 3.7 Flash 1494
  15. GLM-5.3-Flash 1491

Sponsored placements are available on pages like this one. Advertise on Noometry

All results

#ModelProviderRatingSettingSourceDate
1Gemini 4 Argon Google1548highLMArena2026-10-08
2Claude Opus 5.5 Anthropic1535highLMArena2026-10-08
3Claude Opus 4.6 Anthropic1527highLMArena2026-10-08
4Claude Opus 5 Anthropic1526LMArena2026-10-08
5Claude Fable 5.1 Anthropic1526LMArena2026-10-08
6MiMo-V2.6-Pro Xiaomi1512LMArena2026-10-08
7Gemini 3.8 Flash Google1508highLMArena2026-10-08
8Claude Fable 5 Anthropic1508highLMArena2026-10-08
9Claude Opus 4.7 Anthropic1506highLMArena2026-10-08
10Muse Spark 1.3 Meta1503LMArena2026-10-08
11Qwen3.8 Max Alibaba (Qwen)1496LMArena2026-10-08
12Kimi K3 Moonshot AI1496LMArena2026-10-08
13Claude Sonnet 5.5 Anthropic1495xhighLMArena2026-10-08
14Gemini 3.7 Flash Google1494highLMArena2026-10-08
15GLM-5.3-Flash Z.ai (Zhipu)1491LMArena2026-10-08
16GLM-5.3 Z.ai (Zhipu)1489LMArena2026-10-08
17GPT-5.5 OpenAI1489highLMArena2026-10-08
18Gemini 3.5 Flash Google1488highLMArena2026-10-08
19MiMo-V2.5-Pro Xiaomi1488LMArena2026-10-08
20Muse Spark 1.2 Meta1486muse-spark-1.2 (xHigh)LMArena2026-10-08
21Muse Spark 1.1 Meta1486LMArena2026-10-08
22Gemini 3.1 Pro Preview Google1485LMArena2026-10-08
23Gemini 3.6 Flash Google1485highLMArena2026-10-08
24GPT-5.4 OpenAI1485highLMArena2026-10-08
25Claude Sonnet 4.6 Anthropic1484LMArena2026-10-08
26GPT-5.6 Sol OpenAI1484xhighLMArena2026-10-08
27Qwen3.5 Max Preview Alibaba (Qwen)1483LMArena2026-10-08
28Qwen3.7 Max Alibaba (Qwen)1483LMArena2026-10-08
29DeepSeek V4.1 Flash DeepSeek1483LMArena2026-10-08
30MiMo-V2.6-Flash Xiaomi1482LMArena2026-10-08
31Claude Opus 4.8 Anthropic1482highLMArena2026-10-08
32ERNIE 5.1 Baidu1481LMArena2026-10-08
33Gemini 3 Pro Google1480LMArena2026-10-08
34GLM-5.2 Z.ai (Zhipu)1480LMArena2026-10-08
35Claude Opus 4.5 Anthropic1476LMArena2026-10-08
36Muse Spark Meta1474LMArena2026-10-08
37GLM-5.1 Z.ai (Zhipu)1472LMArena2026-10-08
38Kimi K2.6 Moonshot AI1470LMArena2026-10-08
39GPT-5.6 Terra OpenAI1468xhighLMArena2026-10-08
40GPT-6.1 Sol OpenAI1466LMArena2026-10-08
41Gemini 3 Flash Preview Google1465LMArena2026-10-08
42Step 5 Preview StepFun1465Step 5 PreviewLMArena2026-10-08
43Grok 4.5 xAI1462LMArena2026-10-08
44GPT-6 Astra OpenAI1462LMArena2026-10-08
45Claude Sonnet 4.5 Anthropic1462LMArena2026-10-08
46DeepSeek V4 Pro DeepSeek1461LMArena2026-10-08
47Claude Sonnet 5 Anthropic1461highLMArena2026-10-08
48Qwen3.8 27B Alibaba (Qwen)1460LMArena2026-10-08
49Qwen3.7 Plus Alibaba (Qwen)1460LMArena2026-10-08
50Amazon Nova Experimental Chat 26 02 10 Amazon1458LMArena2026-10-08
51MiMo-V2-Pro Xiaomi1457LMArena2026-10-08
52GPT-5.1 OpenAI1457highLMArena2026-10-08
53Qwen3.6 Max Preview Alibaba (Qwen)1457LMArena2026-10-08
54Gemini 2.5 Pro Google1455LMArena2026-10-08
55Kimi K2.5 Moonshot AI1453thinkingLMArena2026-10-08
56Seed 2.0 Pro ByteDance Seed1453LMArena2026-10-08
57Nemotron 3 Ultra NVIDIA1452LMArena2026-10-08
58GLM-5 Z.ai (Zhipu)1452LMArena2026-10-08
59Grok 4.20 (Non-Reasoning) xAI1451LMArena2026-10-08
60GPT-5.6 Luna OpenAI1451xhighLMArena2026-10-08
61Inkling Thinking Machines Lab1451LMArena2026-10-08
62MiMo-V2.5 Xiaomi1450LMArena2026-10-08
63Qwen3.6 Plus Alibaba (Qwen)1449LMArena2026-10-08
64Grok 4.20 Multi-Agent xAI1448LMArena2026-10-08
65Qwen3.5 397B-A17B Alibaba (Qwen)1448LMArena2026-10-08
66Qwen3 Max Alibaba (Qwen)1448LMArena2026-10-08
67Gemma 4 31B IT Google1448LMArena2026-10-08
68Grok 4.6 xAI1447highLMArena2026-10-08
69Hy3 Tencent1447LMArena2026-10-08
70MiniMax-M3 MiniMax1447LMArena2026-10-08
71ERNIE 5.0 0110 Baidu1445LMArena2026-10-08
72GPT-5.2 OpenAI1445LMArena2026-10-08
73MiMo-V2-Omni Xiaomi1445LMArena2026-10-08
74Mistral Large 4 Mistral AI1444LMArena2026-10-08
75DeepSeek V4 Flash DeepSeek1444LMArena2026-10-08
76GLM-5V-Turbo Z.ai (Zhipu)1443LMArena2026-10-08
77Kimi K2.5 Instant Moonshot AI1443LMArena2026-10-08
78GLM-4.7 Z.ai (Zhipu)1443LMArena2026-10-08
79Claude Opus 4.1 Anthropic1443thinking-16kLMArena2026-10-08
80Gemini 3.5 Flash Lite Google1441LMArena2026-10-08
81GLM-4.6 Z.ai (Zhipu)1440LMArena2026-10-08
82Longcat Flash Chat Meituan1440LMArena2026-10-08
83Gemma 4 26B A4B IT Google1439LMArena2026-10-08
84Mistral Medium 3.5 Mistral AI1436LMArena2026-10-08
85Grok 4.1 xAI1435LMArena2026-10-08
86DeepSeek-V3.2-Exp DeepSeek1434LMArena2026-10-08
87Grok 3 xAI1434LMArena2026-10-08
88Qwen3 235B-A22B Alibaba (Qwen)1433LMArena2026-10-08
89Mistral Large 3 Mistral AI1429LMArena2026-10-08
90GLM-4.5 Z.ai (Zhipu)1429LMArena2026-10-08
91Qwen3-VL 235B-A22B Alibaba (Qwen)1428LMArena2026-10-08
92Kimi K2 Thinking Turbo Moonshot AI1428LMArena2026-10-08
93Qwen3-Next 80B-A3B Instruct Alibaba (Qwen)1428LMArena2026-10-08
94Amazon Nova Experimental Chat 12 10 Amazon1427LMArena2026-10-08
95GPT-5.5 Instant OpenAI1426LMArena2026-10-08
96DeepSeek-V3.1-Terminus DeepSeek1426thinkingLMArena2026-10-08
97Mistral Medium Mistral AI1426LMArena2026-10-08
98GPT-5.4 mini OpenAI1424highLMArena2026-10-08
99Chatgpt 4o Latest 20250326 OpenAI1424LMArena2026-10-08
100Inkling-Small Thinking Machines Lab1423Inkling SmallLMArena2026-10-08
101MiniMax-M2.7 MiniMax1422LMArena2026-10-08
102Gemini 2.5 Flash Google1422LMArena2026-10-08
103Amazon Nova Experimental Chat 11 10 Amazon1422LMArena2026-10-08
104Qwen3.5 122B-A10B Alibaba (Qwen)1421LMArena2026-10-08
105Claude Haiku 4.5 Anthropic1420LMArena2026-10-08
106MiMo-V2-Flash Xiaomi1420LMArena2026-10-08
107GPT-6 Sol OpenAI1418LMArena2026-10-08
108DeepSeek-V3.1 DeepSeek1417thinkingLMArena2026-10-08
109Hunyuan Vision 1.5 Tencent1416thinkingLMArena2026-10-08
110GPT-5 OpenAI1416highLMArena2026-10-08
111DeepSeek-R1 DeepSeek1416LMArena2026-10-08
112Amazon Nova Experimental Chat 26 01 10 Amazon1415LMArena2026-10-08
113Qwen3.5 27B Alibaba (Qwen)1414LMArena2026-10-08
114Grok 4.7 xAI1413xhighLMArena2026-10-08
115Grok 4 Fast xAI1412LMArena2026-10-08
116GPT-6 Luna OpenAI1411LMArena2026-10-08
117Step 3.5 Flash StepFun1411LMArena2026-10-08
118MiniMax-M2.1 MiniMax1411LMArena2026-10-08
119Grok 4 xAI1409LMArena2026-10-08
120Gemini 3.1 Flash Lite Google1407LMArena2026-10-08
121Grok 4.1 Fast xAI1407LMArena2026-10-08
122GPT-4.5 OpenAI1403LMArena2026-10-08
123Qwen3.5-Flash Alibaba (Qwen)1403LMArena2026-10-08
124o3 OpenAI1402LMArena2026-10-08
125Qwen3.5 35B-A3B Alibaba (Qwen)1400LMArena2026-10-08
126Claude Opus 4 Anthropic1399thinking-16kLMArena2026-10-08
127GPT-5.3 Chat OpenAI1399LMArena2026-10-08
128Hunyuan T1 20250711 Tencent1399LMArena2026-10-08
129Solar Pro4 Upstage1399LMArena2026-10-08
130Qwen3-30B-A3B Alibaba (Qwen)1398LMArena2026-10-08
131Grok 4.3 xAI1396LMArena2026-10-08
132Muse Glimmer Meta1396LMArena2026-10-08
133Amazon Nova Experimental Chat 10 20 Amazon1396LMArena2026-10-08
134Nemotron 3 Super NVIDIA1386LMArena2026-10-08
135GPT-4.1 OpenAI1384LMArena2026-10-08
136Kimi K2 (Jul 2025) Moonshot AI1384LMArena2026-10-08
137GPT-5.4 nano OpenAI1381highLMArena2026-10-08
138GPT-5 Mini OpenAI1380highLMArena2026-10-08
139GLM-4.5-Air Z.ai (Zhipu)1379LMArena2026-10-08
140Gemini 2.5 Flash-Lite Google1377no-thinkingLMArena2026-10-08
141Grok-3 mini xAI1375highLMArena2026-10-08
142Granite 4.2 30b IBM1374LMArena2026-10-08
143Hunyuan Turbos 20250226 Tencent1374LMArena2026-10-08
144Claude Sonnet 4 Anthropic1372thinking-32kLMArena2026-10-08
145MiniMax-M2.5 MiniMax1372LMArena2026-10-08
146Qwen3-Coder 480B-A35B Instruct Alibaba (Qwen)1372LMArena2026-10-08
147o1 OpenAI1371LMArena2026-10-08
148GLM-4.6V Z.ai (Zhipu)1368LMArena2026-10-08
149o3-mini OpenAI1366highLMArena2026-10-08
150DeepSeek-V3 DeepSeek1365LMArena2026-10-08
151gpt-oss-120b OpenAI1364LMArena2026-10-08
152Nova 2 Lite Amazon1364LMArena2026-10-08
153Mercury 2 Inception1362LMArena2026-10-08
154Qwen2.5-Max Alibaba (Qwen)1360LMArena2026-10-08
155MiniMax-M2 MiniMax1357LMArena2026-10-08
156GLM-4.7-Flash Z.ai (Zhipu)1356LMArena2026-10-08
157Amazon Nova Experimental Chat 10 09 Amazon1356LMArena2026-10-08
158Step 3 StepFun1355LMArena2026-10-08
159o4-mini OpenAI1351LMArena2026-10-08
160Trinity Large Thinking Arcee AI1350LMArena2026-10-08
161GPT-4.1 mini OpenAI1349LMArena2026-10-08
162Nemotron 3 Nano 30B A3B NVIDIA1348LMArena2026-10-08
163Gemini 2.0 Flash (Feb 2025) Google1346LMArena2026-10-08
164Gemma 3 27B Google1340LMArena2026-10-08
165MiniMax M1 MiniMax1339LMArena2026-10-08
166Nemotron 3.5 Lightning NVIDIA1337LMArena2026-10-08
167Nvidia Llama 3.3 Nemotron Super 49b v1.5 NVIDIA1336LMArena2026-10-08
168Mistral Small Mistral AI1335LMArena2026-10-08
169Step 1o Turbo 202506 StepFun1335LMArena2026-10-08
170Qwen3 32B Alibaba (Qwen)1334LMArena2026-10-08
171GLM-4.5V Z.ai (Zhipu)1334LMArena2026-10-08
172Claude 3.7 Sonnet Anthropic1333thinking-32kLMArena2026-10-08
173o1-mini OpenAI1333LMArena2026-10-08
174Granite 4.2 8B IBM1329LMArena2026-10-08
175GPT-5 Nano OpenAI1328highLMArena2026-10-08
176Command A Cohere1326LMArena2026-10-08
177QwQ-32B Alibaba (Qwen)1325LMArena2026-10-08
178Gemini 2.0 Flash-Lite Google1324LMArena2026-10-08
179Olmo 3.1 32b Instruct Allen Institute for AI (Ai2)1322LMArena2026-10-08
180Qwen Plus Alibaba (Qwen)1317LMArena2026-10-08
181Llama 3.1 Nemotron Ultra 253b v1 NVIDIA1316LMArena2026-10-08
182Llama 3.3 Nemotron 49b Super v1 NVIDIA1311LMArena2026-10-08
183Gemma 3 12B Google1309LMArena2026-10-08
184Hunyuan Turbo 0110 Tencent1308LMArena2026-10-08
185Granite 4.2 3b IBM1306LMArena2026-10-08
186Claude 3.5 Sonnet Anthropic1305LMArena2026-10-08
187Yi-Lightning 01.AI1302LMArena2026-10-08
188Olmo 3 32b Think Allen Institute for AI (Ai2)1302LMArena2026-10-08
189Qwen2.5 Plus 1127 Alibaba (Qwen)1299LMArena2026-10-08
190Step 2 16k Exp 202412 StepFun1299LMArena2026-10-08
191Gemini 1.5 Pro (May 2024) Google1296LMArena2026-10-08
192Granite 4.1 8b IBM1293LMArena2026-10-08
193DeepSeek-V2.5 (Sep 2024) DeepSeek1289LMArena2026-10-08
194Molmo 2 8b Allen Institute for AI (Ai2)1287LMArena2026-10-08
195Hunyuan Large 2025 02 10 Tencent1286LMArena2026-10-08
196GPT-4.1 nano OpenAI1286LMArena2026-10-08
197Mercury Inception1285LMArena2026-10-08
198Gemma 3n E4b IT Google1284LMArena2026-10-08
199GPT-4o OpenAI1281LMArena2026-10-08
200Llama 4 Maverick Meta1281LMArena2026-10-08
201Mistral Small 3.1 Mistral AI1278LMArena2026-10-08
202gpt-oss-20b OpenAI1274LMArena2026-10-08
203Olmo 3.1 32b Think Allen Institute for AI (Ai2)1272LMArena2026-10-08
204Grok-2 (Dec 2024) xAI1272LMArena2026-10-08
205Qwen2.5 72B Instruct Alibaba (Qwen)1271LMArena2026-10-08
206Qwen Max Alibaba (Qwen)1269LMArena2026-10-08
207Llama 3.1-405B Meta1269LMArena2026-10-08
208GPT-4o mini OpenAI1267LMArena2026-10-08
209Magistral Medium Mistral AI1267LMArena2026-10-08
210Llama 3.1 Nemotron 70b Instruct NVIDIA1266LMArena2026-10-08
211Llama 4 Scout Meta1266LMArena2026-10-08
212Hunyuan Standard 2025 02 10 Tencent1264LMArena2026-10-08
213Gemini 1.5 Flash (May 2024) Google1257LMArena2026-10-08
214Llama-3.3-70B-Instruct Meta1257LMArena2026-10-08
215Mistral Large Mistral AI1257LMArena2026-10-08
216Hunyuan Large Vision Tencent1257LMArena2026-10-08
217Grok 2 Mini 2024 08 13 xAI1255LMArena2026-10-08
218Gemma 3 4B Google1253LMArena2026-10-08
219Qwen2.5-Coder-32B Alibaba (Qwen)1251LMArena2026-10-08
220Claude 3.5 Haiku Anthropic1251LMArena2026-10-08
221GPT-4 Turbo OpenAI1251LMArena2026-10-08
222Gemini Advanced 0514 Google1248LMArena2026-10-08
223Amazon Nova Pro Amazon1246LMArena2026-10-08
224Claude 3 Opus Anthropic1245LMArena2026-10-08
225GPT-4 OpenAI1241LMArena2026-10-08
226Llama 3.1-70B Meta1241LMArena2026-10-08
227Granite 4.0 H Small IBM1240LMArena2026-10-08
228Mistral Small 3 Mistral AI1233LMArena2026-10-08
229Tulu 3 (Tülu 3) 70B Allen Institute for AI (Ai2)1220LMArena2026-10-08
230Amazon Nova Lite Amazon1220LMArena2026-10-08
231Phi-4 Microsoft1220LMArena2026-10-08
232Gemini 1.5 Flash 8B Google1209LMArena2026-10-08
233Olmo 2 0325 32b Instruct Allen Institute for AI (Ai2)1208LMArena2026-10-08
234Deepseek Coder v2 DeepSeek1207LMArena2026-10-08
235Llama 3.1 Nemotron 51b Instruct NVIDIA1203LMArena2026-10-08
236Nemotron 4 340b Instruct NVIDIA1202LMArena2026-10-08
237Gemma 2 27B Google1198LMArena2026-10-08
238Claude 3 Sonnet Anthropic1197LMArena2026-10-08
239Llama 3-70B Meta1195LMArena2026-10-08
240C4ai Aya Expanse 32b Cohere1193LMArena2026-10-08
241Ministral 8B Mistral AI1191LMArena2026-10-08
242Amazon Nova Micro Amazon1191LMArena2026-10-08
243Qwen2-72B Alibaba (Qwen)1191LMArena2026-10-08
244Command R+ Cohere1186LMArena2026-10-08
245Llama 3.1-8B Meta1175LMArena2026-10-08
246Llama 3.1 Tulu 3 8b Allen Institute for AI (Ai2)1174LMArena2026-10-08
247Claude 3 Haiku Anthropic1174LMArena2026-10-08
248Gemma 2 9B Google1171LMArena2026-10-08
249Qwen1.5-110B Alibaba (Qwen)1168LMArena2026-10-08
250Command R Cohere1164LMArena2026-10-08
251Yi-1.5-34B 01.AI1160LMArena2026-10-08
252C4ai Aya Expanse 8b Cohere1156LMArena2026-10-08
253Mixtral 8x22B Mistral AI1150LMArena2026-10-08
254Mistral Mistral AI1149mediumLMArena2026-10-08
255Qwen1.5-72B Alibaba (Qwen)1148LMArena2026-10-08
256Granite 3.1 8b Instruct IBM1145LMArena2026-10-08
257Granite 3.1 2b Instruct IBM1138LMArena2026-10-08
258Llama 3-8B Meta1133LMArena2026-10-08
259Qwen1.5-32B Alibaba (Qwen)1130LMArena2026-10-08
260Phi 3 Medium 4k Instruct Microsoft1127LMArena2026-10-08
261Mixtral 8x7B Mistral AI1115LMArena2026-10-08
262Qwen1.5-14B Alibaba (Qwen)1113LMArena2026-10-08
263DBRX Databricks1113LMArena2026-10-08
264Gemma 2 2b IT Google1113LMArena2026-10-08
265Gemini 1.0 Pro Google1109LMArena2026-10-08
266GPT-3.5-turbo OpenAI1108LMArena2026-10-08
267Yi-34B 01.AI1104LMArena2026-10-08
268Phi 3 Small 8k Instruct Microsoft1100LMArena2026-10-08
269Llama 3.2 3B Meta1095LMArena2026-10-08
270Granite 3.0 8b Instruct IBM1092LMArena2026-10-08
271Phi 3 Mini 4k Instruct June 2024 Microsoft1087LMArena2026-10-08
272Wizardlm 70b Microsoft1079LMArena2026-10-08
273Granite 3.0 2b Instruct IBM1073LMArena2026-10-08
274Llama 2-70B Meta1073LMArena2026-10-08
275Phi 3 Mini 4k Instruct Microsoft1072LMArena2026-10-08
276Gemma 1.1 7b IT Google1071LMArena2026-10-08
277DeepSeek LLM 67B DeepSeek1070LMArena2026-10-08
278Mistral 7B Mistral AI1067LMArena2026-10-08
279Qwen1.5-7B Alibaba (Qwen)1065LMArena2026-10-08
280Codellama 70b Instruct Meta1052LMArena2026-10-08
281Llama 2-13B Meta1051LMArena2026-10-08
282Llama2 70b Steerlm Chat NVIDIA1047LMArena2026-10-08
283Llama 3.2 1B Meta1044LMArena2026-10-08
284Gemma 7B Google1042LMArena2026-10-08
285Codellama 34b Instruct Meta1032LMArena2026-10-08
286Phi 3 Mini 128k Instruct Microsoft1028LMArena2026-10-08
287Qwen-14B Alibaba (Qwen)1027LMArena2026-10-08
288Wizardlm 13b Microsoft1018LMArena2026-10-08
289Llama 2-7B Meta1009LMArena2026-10-08
290Falcon-180B Technology Innovation Institute1007LMArena2026-10-08
291Gemma 1.1 2b IT Google1005LMArena2026-10-08
292Palm 2 Google1005LMArena2026-10-08
293Olmo 7b Instruct Allen Institute for AI (Ai2)993LMArena2026-10-08
294Gemma 2B Google989LMArena2026-10-08
295Qwen1.5 4b Chat Alibaba (Qwen)976LMArena2026-10-08
296Dolly 2.0-12b Databricks804LMArena2026-10-08
297Llama 13b Meta728LMArena2026-10-08

Compare the leaders

Other reasoning benchmarks

Frequently asked questions

What does LMArena Hard Prompts measure?

LMArena ratings restricted to prompts classified as hard.

Which model has the highest LMArena Hard Prompts score?

As of October 2026, Gemini 4 Argon has the highest published LMArena Hard Prompts score on Noometry at 1548, out of 297 models with results.

What is the best open-weight model on LMArena Hard Prompts?

MiMo-V2.6-Pro has the highest LMArena Hard Prompts rating among open-weight models at 1512, ranking 6 of 297 overall.