Instruction Following benchmark

LMArena Instruction Following leaderboard

As of October 2026, Gemini 4 Argon has the highest published LMArena Instruction Following score on Noometry at 1538, out of 298 models with results.

Last verified

About LMArena Instruction Following

LMArena ratings restricted to prompts with explicit instructions.

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

Top 15 models

Top models on LMArena Instruction Following
  1. Gemini 4 Argon 1538
  2. Claude Opus 5.5 1537
  3. Claude Opus 4.6 1523
  4. Claude Opus 5 1517
  5. Claude Fable 5.1 1517
  6. Claude Fable 5 1502
  7. Claude Opus 4.7 1498
  8. Claude Sonnet 5.5 1495
  9. MiMo-V2.6-Pro 1493
  10. Gemini 3.8 Flash 1490
  11. Kimi K3 1483
  12. Gemini 3.7 Flash 1483
  13. GPT-5.6 Sol 1482
  14. Qwen3.8 Max 1479
  15. GPT-5.5 1479

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

All results

LMArena Instruction Following results by model
#ModelProviderRatingSettingSourceDate
1Gemini 4 Argon Google1538highLMArena2026-10-08
2Claude Opus 5.5 Anthropic1537highLMArena2026-10-08
3Claude Opus 4.6 Anthropic1523highLMArena2026-10-08
4Claude Opus 5 Anthropic1517LMArena2026-10-08
5Claude Fable 5.1 Anthropic1517LMArena2026-10-08
6Claude Fable 5 Anthropic1502highLMArena2026-10-08
7Claude Opus 4.7 Anthropic1498highLMArena2026-10-08
8Claude Sonnet 5.5 Anthropic1495xhighLMArena2026-10-08
9MiMo-V2.6-Pro Xiaomi1493LMArena2026-10-08
10Gemini 3.8 Flash Google1490highLMArena2026-10-08
11Kimi K3 Moonshot AI1483LMArena2026-10-08
12Gemini 3.7 Flash Google1483highLMArena2026-10-08
13GPT-5.6 Sol OpenAI1482xhighLMArena2026-10-08
14Qwen3.8 Max Alibaba (Qwen)1479LMArena2026-10-08
15GPT-5.5 OpenAI1479highLMArena2026-10-08
16Claude Opus 4.5 Anthropic1478LMArena2026-10-08
17GLM-5.3-Flash Z.ai (Zhipu)1478LMArena2026-10-08
18MiMo-V2.5-Pro Xiaomi1477LMArena2026-10-08
19Muse Spark 1.3 Meta1477LMArena2026-10-08
20GLM-5.3 Z.ai (Zhipu)1477LMArena2026-10-08
21Claude Opus 4.8 Anthropic1476highLMArena2026-10-08
22Claude Sonnet 4.6 Anthropic1475LMArena2026-10-08
23DeepSeek V4.1 Flash DeepSeek1474LMArena2026-10-08
24GPT-5.4 OpenAI1469highLMArena2026-10-08
25GPT-6.1 Sol OpenAI1468LMArena2026-10-08
26Gemini 3.5 Flash Google1467highLMArena2026-10-08
27Qwen3.5 Max Preview Alibaba (Qwen)1467LMArena2026-10-08
28Gemini 3.1 Pro Preview Google1466LMArena2026-10-08
29Gemini 3.6 Flash Google1466highLMArena2026-10-08
30GLM-5.2 Z.ai (Zhipu)1465LMArena2026-10-08
31MiMo-V2.6-Flash Xiaomi1463LMArena2026-10-08
32Muse Spark 1.2 Meta1461muse-spark-1.2 (xHigh)LMArena2026-10-08
33ERNIE 5.1 Baidu1460LMArena2026-10-08
34Qwen3.7 Max Alibaba (Qwen)1460LMArena2026-10-08
35Claude Sonnet 4.5 Anthropic1459LMArena2026-10-08
36Gemini 3 Pro Google1458LMArena2026-10-08
37Muse Spark 1.1 Meta1457LMArena2026-10-08
38GPT-5.6 Terra OpenAI1454xhighLMArena2026-10-08
39Claude Sonnet 5 Anthropic1452highLMArena2026-10-08
40GLM-5.1 Z.ai (Zhipu)1451LMArena2026-10-08
41Kimi K2.6 Moonshot AI1451LMArena2026-10-08
42GPT-6 Astra OpenAI1450LMArena2026-10-08
43DeepSeek V4 Pro DeepSeek1448LMArena2026-10-08
44Grok 4.5 xAI1446LMArena2026-10-08
45MiMo-V2-Pro Xiaomi1445LMArena2026-10-08
46Step 5 Preview StepFun1444Step 5 PreviewLMArena2026-10-08
47GPT-5.1 OpenAI1443highLMArena2026-10-08
48Muse Spark Meta1442LMArena2026-10-08
49Qwen3.7 Plus Alibaba (Qwen)1440LMArena2026-10-08
50Qwen3.8 27B Alibaba (Qwen)1439LMArena2026-10-08
51Qwen3.6 Max Preview Alibaba (Qwen)1438LMArena2026-10-08
52Gemini 3 Flash Preview Google1437LMArena2026-10-08
53GPT-5.6 Luna OpenAI1437xhighLMArena2026-10-08
54Gemini 2.5 Pro Google1437LMArena2026-10-08
55Claude Opus 4.1 Anthropic1435thinking-16kLMArena2026-10-08
56MiMo-V2.5 Xiaomi1434LMArena2026-10-08
57Gemma 4 31B IT Google1433LMArena2026-10-08
58MiniMax-M3 MiniMax1433LMArena2026-10-08
59Grok 4.6 xAI1431highLMArena2026-10-08
60Kimi K2.5 Moonshot AI1431thinkingLMArena2026-10-08
61Kimi K2.5 Instant Moonshot AI1430LMArena2026-10-08
62MiMo-V2-Omni Xiaomi1428LMArena2026-10-08
63GLM-5 Z.ai (Zhipu)1428LMArena2026-10-08
64Amazon Nova Experimental Chat 26 02 10 Amazon1427LMArena2026-10-08
65Hy3 Tencent1426LMArena2026-10-08
66Inkling Thinking Machines Lab1426LMArena2026-10-08
67Qwen3.6 Plus Alibaba (Qwen)1425LMArena2026-10-08
68Mistral Large 4 Mistral AI1424LMArena2026-10-08
69Qwen3.5 397B-A17B Alibaba (Qwen)1424LMArena2026-10-08
70GLM-5V-Turbo Z.ai (Zhipu)1423LMArena2026-10-08
71DeepSeek V4 Flash DeepSeek1421LMArena2026-10-08
72Gemma 4 26B A4B IT Google1420LMArena2026-10-08
73Grok 4.20 (Non-Reasoning) xAI1420LMArena2026-10-08
74Grok 4.20 Multi-Agent xAI1420LMArena2026-10-08
75Gemini 3.5 Flash Lite Google1420LMArena2026-10-08
76Qwen3 Max Alibaba (Qwen)1419LMArena2026-10-08
77Nemotron 3 Ultra NVIDIA1418LMArena2026-10-08
78GPT-5.2 OpenAI1417LMArena2026-10-08
79Mistral Medium 3.5 Mistral AI1415LMArena2026-10-08
80Claude Haiku 4.5 Anthropic1414LMArena2026-10-08
81Seed 2.0 Pro ByteDance Seed1414LMArena2026-10-08
82ERNIE 5.0 0110 Baidu1413LMArena2026-10-08
83DeepSeek-V3.2-Exp DeepSeek1413LMArena2026-10-08
84GPT-6 Sol OpenAI1412LMArena2026-10-08
85GLM-4.7 Z.ai (Zhipu)1411LMArena2026-10-08
86Longcat Flash Chat Meituan1411LMArena2026-10-08
87GLM-4.6 Z.ai (Zhipu)1410LMArena2026-10-08
88GPT-6 Luna OpenAI1409LMArena2026-10-08
89Grok 3 xAI1409LMArena2026-10-08
90Qwen3 235B-A22B Alibaba (Qwen)1408LMArena2026-10-08
91Claude Opus 4 Anthropic1406thinking-16kLMArena2026-10-08
92GPT-5.5 Instant OpenAI1406LMArena2026-10-08
93Qwen3-VL 235B-A22B Alibaba (Qwen)1406LMArena2026-10-08
94GPT-5.4 mini OpenAI1405highLMArena2026-10-08
95Gemini 2.5 Flash Google1405LMArena2026-10-08
96MiniMax-M2.7 MiniMax1405LMArena2026-10-08
97Grok 4.7 xAI1404xhighLMArena2026-10-08
98GLM-4.5 Z.ai (Zhipu)1404LMArena2026-10-08
99DeepSeek-V3.1-Terminus DeepSeek1404thinkingLMArena2026-10-08
100GPT-4.5 OpenAI1404LMArena2026-10-08
101Chatgpt 4o Latest 20250326 OpenAI1403LMArena2026-10-08
102Mistral Large 3 Mistral AI1403LMArena2026-10-08
103Kimi K2 Thinking Turbo Moonshot AI1403LMArena2026-10-08
104DeepSeek-V3.1 DeepSeek1400thinkingLMArena2026-10-08
105MiniMax-M2.1 MiniMax1400LMArena2026-10-08
106Grok 4.1 xAI1400LMArena2026-10-08
107Inkling-Small Thinking Machines Lab1399Inkling SmallLMArena2026-10-08
108Qwen3.5 122B-A10B Alibaba (Qwen)1399LMArena2026-10-08
109Mistral Medium Mistral AI1398LMArena2026-10-08
110Hunyuan Vision 1.5 Tencent1397thinkingLMArena2026-10-08
111Qwen3.5 27B Alibaba (Qwen)1393LMArena2026-10-08
112MiMo-V2-Flash Xiaomi1392LMArena2026-10-08
113Qwen3-Next 80B-A3B Instruct Alibaba (Qwen)1389LMArena2026-10-08
114GPT-5 OpenAI1388highLMArena2026-10-08
115Grok 4 Fast xAI1387LMArena2026-10-08
116Grok 4 xAI1387LMArena2026-10-08
117Amazon Nova Experimental Chat 11 10 Amazon1387LMArena2026-10-08
118Amazon Nova Experimental Chat 12 10 Amazon1387LMArena2026-10-08
119Step 3.5 Flash StepFun1385LMArena2026-10-08
120DeepSeek-R1 DeepSeek1382LMArena2026-10-08
121Amazon Nova Experimental Chat 26 01 10 Amazon1381LMArena2026-10-08
122Qwen3.5 35B-A3B Alibaba (Qwen)1379LMArena2026-10-08
123GPT-5.3 Chat OpenAI1378LMArena2026-10-08
124Gemini 3.1 Flash Lite Google1377LMArena2026-10-08
125Solar Pro4 Upstage1377LMArena2026-10-08
126Claude Sonnet 4 Anthropic1376thinking-32kLMArena2026-10-08
127Grok 4.1 Fast xAI1376LMArena2026-10-08
128Muse Glimmer Meta1375LMArena2026-10-08
129Hunyuan T1 20250711 Tencent1374LMArena2026-10-08
130Qwen3.5-Flash Alibaba (Qwen)1374LMArena2026-10-08
131o3 OpenAI1368LMArena2026-10-08
132o1 OpenAI1367LMArena2026-10-08
133GPT-4.1 OpenAI1367LMArena2026-10-08
134Gemini 2.5 Flash-Lite Google1367thinkingLMArena2026-10-08
135Grok 4.3 xAI1366LMArena2026-10-08
136Amazon Nova Experimental Chat 10 20 Amazon1365LMArena2026-10-08
137Qwen3-30B-A3B Alibaba (Qwen)1363LMArena2026-10-08
138GPT-5.4 nano OpenAI1362highLMArena2026-10-08
139GPT-5 Mini OpenAI1357highLMArena2026-10-08
140Grok-3 mini xAI1357highLMArena2026-10-08
141Qwen3-Coder 480B-A35B Instruct Alibaba (Qwen)1355LMArena2026-10-08
142GLM-4.5-Air Z.ai (Zhipu)1354LMArena2026-10-08
143MiniMax-M2.5 MiniMax1353LMArena2026-10-08
144Claude 3.7 Sonnet Anthropic1352thinking-32kLMArena2026-10-08
145GLM-4.6V Z.ai (Zhipu)1352LMArena2026-10-08
146Kimi K2 (Jul 2025) Moonshot AI1348LMArena2026-10-08
147Nemotron 3 Super NVIDIA1347LMArena2026-10-08
148Granite 4.2 30b IBM1347LMArena2026-10-08
149DeepSeek-V3 DeepSeek1345LMArena2026-10-08
150Hunyuan Turbos 20250226 Tencent1344LMArena2026-10-08
151o3-mini OpenAI1337highLMArena2026-10-08
152Gemini 2.0 Flash (Feb 2025) Google1336LMArena2026-10-08
153Qwen2.5-Max Alibaba (Qwen)1335LMArena2026-10-08
154Nova 2 Lite Amazon1335LMArena2026-10-08
155Trinity Large Thinking Arcee AI1334LMArena2026-10-08
156GPT-4.1 mini OpenAI1333LMArena2026-10-08
157Step 3 StepFun1332LMArena2026-10-08
158Mercury 2 Inception1329LMArena2026-10-08
159MiniMax-M2 MiniMax1328LMArena2026-10-08
160GLM-4.7-Flash Z.ai (Zhipu)1327LMArena2026-10-08
161Gemma 3 27B Google1321LMArena2026-10-08
162o4-mini OpenAI1321LMArena2026-10-08
163Nemotron 3.5 Lightning NVIDIA1318LMArena2026-10-08
164gpt-oss-120b OpenAI1318LMArena2026-10-08
165Amazon Nova Experimental Chat 10 09 Amazon1315LMArena2026-10-08
166MiniMax M1 MiniMax1312LMArena2026-10-08
167GLM-4.5V Z.ai (Zhipu)1311LMArena2026-10-08
168Mistral Small Mistral AI1310LMArena2026-10-08
169Step 1o Turbo 202506 StepFun1310LMArena2026-10-08
170Command A Cohere1309LMArena2026-10-08
171Llama 3.1 Nemotron Ultra 253b v1 NVIDIA1308LMArena2026-10-08
172GPT-5 Nano OpenAI1306highLMArena2026-10-08
173Qwen3 32B Alibaba (Qwen)1305LMArena2026-10-08
174Gemini 2.0 Flash-Lite Google1305LMArena2026-10-08
175o1-mini OpenAI1304LMArena2026-10-08
176Nemotron 3 Nano 30B A3B NVIDIA1304LMArena2026-10-08
177Qwen Plus Alibaba (Qwen)1303LMArena2026-10-08
178Granite 4.2 8B IBM1301LMArena2026-10-08
179Gemma 3 12B Google1299LMArena2026-10-08
180Olmo 3.1 32b Instruct Allen Institute for AI (Ai2)1299LMArena2026-10-08
181Nvidia Llama 3.3 Nemotron Super 49b v1.5 NVIDIA1299LMArena2026-10-08
182Claude 3.5 Sonnet Anthropic1297LMArena2026-10-08
183QwQ-32B Alibaba (Qwen)1297LMArena2026-10-08
184Gemini 1.5 Pro (May 2024) Google1297LMArena2026-10-08
185Llama 3.3 Nemotron 49b Super v1 NVIDIA1293LMArena2026-10-08
186Step 2 16k Exp 202412 StepFun1287LMArena2026-10-08
187DeepSeek-V2.5 (Sep 2024) DeepSeek1280LMArena2026-10-08
188Hunyuan Turbo 0110 Tencent1278LMArena2026-10-08
189Yi-Lightning 01.AI1278LMArena2026-10-08
190GPT-4o OpenAI1278LMArena2026-10-08
191Hunyuan Large 2025 02 10 Tencent1277LMArena2026-10-08
192Olmo 3 32b Think Allen Institute for AI (Ai2)1275LMArena2026-10-08
193Qwen2.5 Plus 1127 Alibaba (Qwen)1275LMArena2026-10-08
194Granite 4.2 3b IBM1273LMArena2026-10-08
195Molmo 2 8b Allen Institute for AI (Ai2)1270LMArena2026-10-08
196Grok-2 (Dec 2024) xAI1270LMArena2026-10-08
197Granite 4.1 8b IBM1269LMArena2026-10-08
198Llama 4 Maverick Meta1267LMArena2026-10-08
199GPT-4.1 nano OpenAI1267LMArena2026-10-08
200Mistral Small 3.1 Mistral AI1264LMArena2026-10-08
201Qwen Max Alibaba (Qwen)1262LMArena2026-10-08
202Llama 3.1-405B Meta1259LMArena2026-10-08
203Llama 3.1-405B Meta1259LMArena2026-10-08
204GPT-4o mini OpenAI1258LMArena2026-10-08
205Gemini 1.5 Flash (May 2024) Google1258LMArena2026-10-08
206Gemma 3n E4b IT Google1255LMArena2026-10-08
207Qwen2.5 72B Instruct Alibaba (Qwen)1254LMArena2026-10-08
208Magistral Medium Mistral AI1254LMArena2026-10-08
209Gemini Advanced 0514 Google1253LMArena2026-10-08
210Llama 3.1 Nemotron 70b Instruct NVIDIA1252LMArena2026-10-08
211Hunyuan Large Vision Tencent1251LMArena2026-10-08
212GPT-4 Turbo OpenAI1249LMArena2026-10-08
213Mistral Large Mistral AI1249LMArena2026-10-08
214Claude 3 Opus Anthropic1248LMArena2026-10-08
215Llama 4 Scout Meta1248LMArena2026-10-08
216Olmo 3.1 32b Think Allen Institute for AI (Ai2)1247LMArena2026-10-08
217Hunyuan Standard 2025 02 10 Tencent1245LMArena2026-10-08
218Grok 2 Mini 2024 08 13 xAI1245LMArena2026-10-08
219Llama-3.3-70B-Instruct Meta1242LMArena2026-10-08
220Claude 3.5 Haiku Anthropic1241LMArena2026-10-08
221GPT-4 OpenAI1241LMArena2026-10-08
222Mercury Inception1239LMArena2026-10-08
223Gemma 3 4B Google1239LMArena2026-10-08
224gpt-oss-20b OpenAI1236LMArena2026-10-08
225Amazon Nova Pro Amazon1235LMArena2026-10-08
226Tulu 3 (Tülu 3) 70B Allen Institute for AI (Ai2)1233LMArena2026-10-08
227Llama 3.1-70B Meta1231LMArena2026-10-08
228Qwen2.5-Coder-32B Alibaba (Qwen)1223LMArena2026-10-08
229Granite 4.0 H Small IBM1222LMArena2026-10-08
230Mistral Small 3 Mistral AI1214LMArena2026-10-08
231Gemma 2 27B Google1206LMArena2026-10-08
232Amazon Nova Lite Amazon1205LMArena2026-10-08
233Nemotron 4 340b Instruct NVIDIA1203LMArena2026-10-08
234Phi-4 Microsoft1201LMArena2026-10-08
235Llama 3.1 Nemotron 51b Instruct NVIDIA1201LMArena2026-10-08
236Claude 3 Sonnet Anthropic1199LMArena2026-10-08
237Gemini 1.5 Flash 8B Google1199LMArena2026-10-08
238Command R+ Cohere1197LMArena2026-10-08
239C4ai Aya Expanse 32b Cohere1196LMArena2026-10-08
240Llama 3-70B Meta1194LMArena2026-10-08
241Olmo 2 0325 32b Instruct Allen Institute for AI (Ai2)1186LMArena2026-10-08
242Qwen2-72B Alibaba (Qwen)1181LMArena2026-10-08
243Deepseek Coder v2 DeepSeek1180LMArena2026-10-08
244Gemma 2 9B Google1178LMArena2026-10-08
245Llama 3.1 Tulu 3 8b Allen Institute for AI (Ai2)1174LMArena2026-10-08
246Amazon Nova Micro Amazon1174LMArena2026-10-08
247Claude 3 Haiku Anthropic1173LMArena2026-10-08
248Command R Cohere1167LMArena2026-10-08
249Ministral 8B Mistral AI1161LMArena2026-10-08
250Llama 3.1-8B Meta1159LMArena2026-10-08
251Qwen1.5-110B Alibaba (Qwen)1158LMArena2026-10-08
252C4ai Aya Expanse 8b Cohere1155LMArena2026-10-08
253Mistral Mistral AI1152mediumLMArena2026-10-08
254Mixtral 8x22B Mistral AI1147LMArena2026-10-08
255Qwen1.5-72B Alibaba (Qwen)1141LMArena2026-10-08
256Yi-1.5-34B 01.AI1139LMArena2026-10-08
257Granite 3.1 8b Instruct IBM1131LMArena2026-10-08
258Llama 3-8B Meta1127LMArena2026-10-08
259GPT-3.5-turbo OpenAI1119LMArena2026-10-08
260Gemma 2 2b IT Google1118LMArena2026-10-08
261Granite 3.1 2b Instruct IBM1116LMArena2026-10-08
262Qwen1.5-32B Alibaba (Qwen)1116LMArena2026-10-08
263Gemini 1.0 Pro Google1114LMArena2026-10-08
264Phi 3 Medium 4k Instruct Microsoft1113LMArena2026-10-08
265DBRX Databricks1112LMArena2026-10-08
266Mixtral 8x7B Mistral AI1109LMArena2026-10-08
267Qwen1.5-14B Alibaba (Qwen)1102LMArena2026-10-08
268Wizardlm 70b Microsoft1093LMArena2026-10-08
269Yi-34B 01.AI1091LMArena2026-10-08
270Llama 3.2 3B Meta1089LMArena2026-10-08
271Granite 3.0 8b Instruct IBM1088LMArena2026-10-08
272Phi 3 Small 8k Instruct Microsoft1087LMArena2026-10-08
273DeepSeek LLM 67B DeepSeek1079LMArena2026-10-08
274Llama 2-70B Meta1071LMArena2026-10-08
275Llama2 70b Steerlm Chat NVIDIA1060LMArena2026-10-08
276Mistral 7B Mistral AI1060LMArena2026-10-08
277Qwen1.5-7B Alibaba (Qwen)1058LMArena2026-10-08
278Phi 3 Mini 4k Instruct June 2024 Microsoft1058LMArena2026-10-08
279Gemma 1.1 7b IT Google1057LMArena2026-10-08
280Granite 3.0 2b Instruct IBM1056LMArena2026-10-08
281Phi 3 Mini 4k Instruct Microsoft1053LMArena2026-10-08
282Wizardlm 13b Microsoft1048LMArena2026-10-08
283Falcon-180B Technology Innovation Institute1047LMArena2026-10-08
284Llama 2-13B Meta1045LMArena2026-10-08
285Qwen-14B Alibaba (Qwen)1031LMArena2026-10-08
286Llama 3.2 1B Meta1031LMArena2026-10-08
287Codellama 34b Instruct Meta1028LMArena2026-10-08
288Codellama 70b Instruct Meta1024LMArena2026-10-08
289Phi 3 Mini 128k Instruct Microsoft1023LMArena2026-10-08
290Gemma 7B Google1017LMArena2026-10-08
291Palm 2 Google1010LMArena2026-10-08
292Llama 2-7B Meta1006LMArena2026-10-08
293Gemma 1.1 2b IT Google992LMArena2026-10-08
294Qwen1.5 4b Chat Alibaba (Qwen)978LMArena2026-10-08
295Olmo 7b Instruct Allen Institute for AI (Ai2)978LMArena2026-10-08
296Gemma 2B Google970LMArena2026-10-08
297Dolly 2.0-12b Databricks814LMArena2026-10-08
298Llama 13b Meta781LMArena2026-10-08

Compare the leaders

Other instruction following benchmarks

Frequently asked questions

What does LMArena Instruction Following measure?

LMArena ratings restricted to prompts with explicit instructions.

Which model has the highest LMArena Instruction Following score?

As of October 2026, Gemini 4 Argon has the highest published LMArena Instruction Following score on Noometry at 1538, out of 298 models with results.

What is the best open-weight model on LMArena Instruction Following?

MiMo-V2.6-Pro has the highest LMArena Instruction Following rating among open-weight models at 1493, ranking 9 of 298 overall.