LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

LMArena Text: hard prompts leaderboard

Gemini 4 Argon leads
1546
models measured
156
lead over #2
+10 Elo
median
1424

Leader on LMArena Text: hard prompts · read October 21

1 Gemini 4 Argon1546

On October 21, 2026, 156 of our catalog's models had a LMArena Text: hard prompts result, and Google Gemini 4 Argon led with 1546.

Prompts the arena classes as hard: specific, complex, needing problem solving. People ask a question, compare two anonymous answers and vote for the better one; the rating is built from those votes, as LMArena publishes it.

Role in the index

We collect this benchmark and show it as supporting evidence. It is not part of the Vibe Coding Index, because it measures something other than coding and the index is about coding. Read the methodology

156 models with a result

Original board: LMArena (Hugging Face dataset)
LMArena Text: hard prompts results for models in the catalog
# Model Result VotesInterval (±) Evidence Coding index Output $/Mtok
1 Gemini 4 ArgonGoogle 1546 3,131±10.9 80.5 $10.00
2 Claude Opus 5.5Anthropic 1536 5,917±8.5 92.2 $20.00
3 Claude Fable 5.1Anthropic 1528 9,089±7.1 89.1 $50.00
4 Claude Opus 5Anthropic 1527 21,745±5.5 81.8 $25.00
5 Claude Opus 4.6Anthropic 1525 52,600±4.0 66.9 $25.00
6 MiMo-V2.6-ProXiaomi 1511 2,619±11.5 72.6 $0.87
7 Gemini 3.8 FlashGoogle 1509 26,670±5.4 70.5 $4.125
8 Claude Fable 5Anthropic 1507 28,193±4.9 82.1 $50.00
9 Muse WonderMeta 1506 3,187±10.7 64.5 $20.00
10 Claude Opus 4.7Anthropic 1506 43,676±4.4 51.1 $25.00
11 Muse Spark 1.3Meta 1503 12,510±6.4 79.7 $4.25
12 Qwen3.8 Max (0902)Qwen 1496 19,425±5.5 63.5 $6.00
13 Gemini 3.7 FlashGoogle 1495 19,036±5.6 69.9 $3.75
14 Kimi K3Moonshot AI 1495 21,953±5.2 61.8 $11.936
15 Claude Sonnet 5.5Anthropic 1494 5,015±8.9 88.9 $10.00
16 GLM 5.3 FlashZ.ai 1490 22,985±5.5 71.0 $0.50
17 GLM 5.3Z.ai 1490 15,149±6.0 73.1 $1.40
18 Muse Spark 1.2Meta 1489 6,116±8.2 65.1 $4.25
19 GPT-5.5OpenAI 1489 45,527±4.3 67.8 $30.00
20 Gemini 3.5 FlashGoogle 1488 32,389±4.5 42.8 $9.00
21 Gemini 3.6 FlashGoogle 1486 24,133±5.1 48.3 $3.75
22 MiMo-V2.5-ProXiaomi 1486 50,452±4.1 40.0 $0.609
23 Muse Spark 1.1Meta 1485 29,563±4.8 48.9 $4.25
24 GPT-5.4OpenAI 1485 42,124±4.5 37.7 $15.00
25 Gemini 3.1 Pro PreviewGoogle 1485 84,197±3.5 35.3 $12.00
26 Claude Sonnet 4.6Anthropic 1485 46,227±4.2 57.8 $15.00
27 GPT-5.6 SolOpenAI 1483 27,137±5.0 76.7 $20.00
28 DeepSeek V4.1 FlashDeepSeek 1483 8,655±7.1 77.2 $0.525
29 Claude Opus 4.8Anthropic 1482 42,516±4.5 66.0 $25.00
30 GLM 5.2Z.ai 1481 33,290±4.7 57.7 $7.00
31 Claude Opus 4.5Anthropic 1476 43,232±3.9 66.8 $25.00
32 MiMo-V2.6-FlashXiaomi 1475 7,817±7.4 61.9 $0.28
33 GLM 5.1Z.ai 1473 42,115±4.1 44.0 $4.40
34 Kimi K2.6Moonshot AI 1470 26,060±5.2 44.2 $3.41
35 GPT-5.6 TerraOpenAI 1469 29,450±4.9 69.0 $12.00
36 GPT-6 AstraOpenAI 1463 8,080±7.3 88.7 $50.00
37 GPT-6.1 SolOpenAI 1463 4,318±9.6 88.7 $10.00
38 Claude Sonnet 4.5Anthropic 1462 48,301±3.6 59.1 $15.00
39 Grok 4.5xAI 1462 26,637±5.0 75.5 $6.00
40 Step 5 PreviewStepFun 1461 5,502±8.4 71.4 $2.70
41 DeepSeek V4 Pro 0423DeepSeek 1461 38,504±4.5 35.3 $1.9105
42 Claude Sonnet 5Anthropic 1461 33,444±4.7 63.5 $10.00
43 Qwen3.7 PlusQwen 1460 28,278±5.0 Not benchmarked $1.28
44 Qwen3.8 27BQwen 1459 16,357±5.8 73.6 $1.35
45 DeepSeek V4 Pro 0813DeepSeek 1458 6,526±7.7 54.6 $1.98
46 GPT-5.1OpenAI 1457 22,416±4.8 53.3 $10.00
47 Gemini 2.5 ProGoogle 1455 65,892±3.1 37.0 $10.00
48 Kimi K2.5Moonshot AI 1453 47,893±3.9 58.5 $2.50
49 Nemotron 3 UltraNVIDIA 1452 7,587±7.7 Not benchmarked $2.40
50 GLM 5Z.ai 1452 18,205±5.1 59.7 $3.20
51 GPT-5.6 LunaOpenAI 1451 26,984±5.0 61.5 $1.20
52 T InklingThinking Machines 1451 27,837±5.0 49.7 $4.05
53 MiMo-V2.5Xiaomi 1450 31,827±4.9 34.0 $0.238
54 Qwen3.6 PlusQwen 1448 32,549±4.8 37.4 $1.95
55 Grok 4.6xAI 1448 20,007±5.5 76.4 $6.00
56 Qwen3.5 397B A17BQwen 1448 62,653±3.6 28.6 $2.34
57 Gemma 4 31BGoogle 1448 3,565±9.6 38.3 $0.40
58 Hy3Tencent 1447 8,479±7.1 43.9 $0.528
59 MiniMax M3MiniMax 1447 43,582±4.4 57.8 $1.20
60 Mistral Large 4Mistral AI 1444 5,844±8.2 55.3 $4.18
61 DeepSeek V4 Flash 0423DeepSeek 1444 34,751±4.7 54.6 $0.182
62 GLM 5V TurboZ.ai 1443 6,952±7.6 28.8 $4.00
63 GLM 4.7Z.ai 1443 6,793±7.7 33.6 $1.98
64 Claude Opus 4.1Anthropic 1442 24,961±4.5 27.2 $75.00
65 R Beam 1.0 501B A23BReflection 1442 2,676±11.6 33.8 $1.00
66 Gemini 3.5 Flash LiteGoogle 1441 23,808±5.2 33.6 $2.50
67 GLM 4.6Z.ai 1440 19,444±5.0 52.2 $2.20
68 Gemma 4 26B A4BGoogle 1438 3,471±9.7 22.8 $0.255
69 Mistral Medium 3.5Mistral AI 1436 7,650±7.7 9.2 $7.50
70 DeepSeek V3.2DeepSeek 1434 27,069±4.4 54.4 $0.42
71 GLM 4.5Z.ai 1429 11,114±6.4 64.2 $2.20
72 DeepSeek V3.2 ExpDeepSeek 1429 6,580±7.9 55.0 $0.41
73 Qwen3 VL 235B A22B InstructQwen 1428 5,704±8.6 Not benchmarked $1.90
74 GPT-5.2OpenAI 1428 29,881±4.4 58.9 $14.00
75 Qwen3 Next 80B A3B InstructQwen 1428 11,998±6.0 Not benchmarked $1.10
76 Mistral Medium 3.1Mistral AI 1426 52,905±3.3 Not benchmarked $2.00
77 DeepSeek V3.1 TerminusDeepSeek 1426 1,659±14.3 Not benchmarked $1.00
78 GPT-5.4 MiniOpenAI 1424 41,648±4.4 28.2 $4.50
79 T Inkling SmallThinking Machines 1423 22,270±5.3 29.9 $1.20
80 MiniMax M2.7MiniMax 1422 56,288±3.9 28.4 $1.20
81 Gemini 2.5 FlashGoogle 1422 65,128±3.1 28.7 $2.50
82 GPT-6 SolOpenAI 1422 8,512±7.3 82.8 $10.00
83 Qwen3.5-122B-A10BQwen 1421 18,809±5.2 22.9 $2.08
84 Claude Haiku 4.5Anthropic 1421 93,248±3.0 49.8 $5.00
85 Qwen3 235B A22B Thinking 2507Qwen 1416 3,806±9.5 Not benchmarked $3.50
86 R1 0528DeepSeek 1415 6,879±7.7 71.4 $2.15
87 GPT-5OpenAI 1415 15,055±5.7 72.2 $10.00
88 Grok 4.7xAI 1415 8,230±7.4 75.0 $6.00
89 Qwen3.5-27BQwen 1415 18,079±5.2 22.6 $2.16
90 Hy3 previewTencent 1413 4,589±9.1 22.4 $0.60
91 Step 3.5 FlashStepFun 1411 37,536±4.3 Not benchmarked $0.30
92 GPT-6 LunaOpenAI 1410 39,234±5.1 71.4 $0.50
93 Gemini 3.1 Flash Lite PreviewGoogle 1407 41,605±4.4 8.1 $1.50
94 Qwen3.5-FlashQwen 1403 40,118±4.4 6.2 $0.26
95 o3OpenAI 1402 25,528±4.8 70.7 $8.00
96 Qwen3.5-35B-A3BQwen 1400 19,346±5.1 7.7 $1.00
97 Solar Pro 4Upstage 1400 15,023±6.0 24.4 $0.60
98 Qwen3 30B A3B Instruct 2507Qwen 1398 10,974±6.3 Not benchmarked $0.30
99 Grok 4.3xAI 1396 47,665±4.3 22.5 $2.50
100 Nemotron 3 SuperNVIDIA 1386 4,396±9.0 Not benchmarked $2.40
101 GPT-4.1OpenAI 1384 22,016±4.9 46.5 $8.00
102 GPT-5.4 NanoOpenAI 1381 41,305±4.5 Not benchmarked $1.25
103 GPT-5 MiniOpenAI 1380 12,861±6.0 59.8 $2.00
104 GLM 4.5 AirZ.ai 1379 14,897±5.5 Not benchmarked $0.85
105 MiniMax M2.5MiniMax 1372 26,855±4.8 59.4 $0.95
106 Qwen3 Coder 480B A35BQwen 1372 11,532±6.4 50.0 $1.00
107 o1OpenAI 1371 6,453±7.7 61.7 $60.00
108 Qwen3 Next 80B A3B ThinkingQwen 1371 6,666±7.5 Not benchmarked $1.20
109 GLM 4.6VZ.ai 1367 1,532±14.9 Not benchmarked $0.90
110 o3 MiniOpenAI 1365 4,386±9.1 52.1 $4.40
111 Mistral Medium 3Mistral AI 1364 13,652±6.0 Not benchmarked $2.00
112 Nova 2 LiteAmazon 1364 6,632±7.8 Not benchmarked $2.50
113 gpt-oss-120bOpenAI 1364 14,687±5.6 34.5 $0.60
114 Mercury 2Inception 1362 1,844±13.3 0.0 $0.75
115 MiniMax M2MiniMax 1357 3,748±9.8 45.3 $1.02
116 GLM 4.7 FlashZ.ai 1356 6,739±7.3 Not benchmarked $0.40
117 o4 MiniOpenAI 1351 19,383±5.1 68.6 $4.40
118 GPT-4.1 MiniOpenAI 1349 16,063±5.6 28.5 $1.60
119 Nemotron 3 Nano 30B A3BNVIDIA 1348 8,383±7.0 Not benchmarked $0.24
120 Trinity Large ThinkingArcee AI 1342 20,054±5.6 5.4 $0.80
121 Gemma 3 27BGoogle 1340 17,704±5.1 4.9 $0.35
122 MiniMax M1MiniMax 1339 15,596±5.5 Not benchmarked $2.20
123 GLM 4.5VZ.ai 1335 2,346±12.0 Not benchmarked $1.80
124 Qwen3 32BQwen 1334 1,207±15.9 40.0 $0.28
125 GPT-5 NanoOpenAI 1328 3,782±9.7 34.8 $0.40
126 Granite 4.2 8BIBM Granite 1327 2,076±13.9 Not benchmarked $0.25
127 Command ACohere 1326 23,764±4.6 Not benchmarked $10.00
128 Qwen-PlusQwen 1317 1,481±14.1 Not benchmarked $0.78
129 Qwen3 30B A3BQwen 1314 10,646±6.3 Not benchmarked $0.50
130 Gemma 3 12BGoogle 1309 977±17.6 Not benchmarked $0.15
131 GPT-4.1 NanoOpenAI 1286 1,669±13.5 8.9 $0.40
132 GPT-4o (2024-05-13)OpenAI 1281 32,416±5.3 Not benchmarked $15.00
133 Llama 4 MaverickMeta Llama 1281 15,865±5.7 18.1 $0.6525
134 Mistral Small 3.1 24BMistral AI 1278 14,623±5.8 Not benchmarked $0.555
135 gpt-oss-20bOpenAI 1274 4,719±8.9 Not benchmarked $0.09
136 Qwen2.5 72B InstructQwen 1271 10,545±6.3 Not benchmarked $0.40
137 GPT-4o-mini (2024-07-18)OpenAI 1267 18,259±5.3 3.6 $0.60
138 Llama 4 ScoutMeta Llama 1266 12,899±6.2 9.1 $0.30
139 GPT-4o (2024-08-06)OpenAI 1264 12,655±6.4 23.1 $10.00
140 Llama 3.3 70B InstructMeta Llama 1257 16,963±5.0 Not benchmarked $0.32
141 Mistral LargeMistral AI 1257 6,954±7.0 Not benchmarked $6.00
142 Gemma 3 4BGoogle 1253 1,092±17.0 Not benchmarked $0.05
143 Qwen2.5 Coder 32B InstructQwen 1251 1,386±14.3 30.5 $1.00
144 GPT-4 TurboOpenAI 1251 28,310±5.9 Not benchmarked $30.00
145 Nova Pro 1.0Amazon 1246 6,300±7.3 Not benchmarked $3.20
146 Llama 3.1 70B InstructMeta Llama 1241 14,981±5.8 Not benchmarked $0.40
147 Mistral Small 3Mistral AI 1233 3,569±9.6 Not benchmarked $0.08
148 Nova Lite 1.0Amazon 1220 4,913±8.3 Not benchmarked $0.24
149 Phi 4Microsoft 1220 5,747±7.8 Not benchmarked $0.14
150 GPT-4OpenAI 1200 13,935±7.7 22.4 $60.00
151 Gemma 2 27BGoogle 1198 20,494±5.2 Not benchmarked $0.65
152 Nova Micro 1.0Amazon 1191 4,859±8.3 Not benchmarked $0.14
153 Command R (08-2024)Cohere 1186 2,822±10.5 Not benchmarked $0.60
154 Llama 3.1 8B InstructMeta Llama 1175 13,535±6.1 Not benchmarked $0.08
155 Llama 3.2 3B InstructMeta Llama 1095 2,216±12.1 Not benchmarked $0.335
156 Llama 3.2 1B InstructMeta Llama 1044 2,247±12.5 Not benchmarked $0.201

Only models in our catalog are listed, and each appears once with its best published run. 172 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out. Every other column is a figure the board publishes for the same run.