LMArena Text: math leaderboard
- Gemini 4 Argon leads
- 1536
- models measured
- 153
- lead over #2
- +3 Elo
- median
- 1425
Leader on LMArena Text: math · read October 21
On October 21, 2026, 153 of our catalog's models had a LMArena Text: math result, and Google Gemini 4 Argon led with 1536.
Prompts that ask for math. 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
| # | Model | Result | Votes | Interval (±) | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|---|---|
| 1 | Gemini 4 ArgonGoogle | 1536 | 273 | ±34.3 | 80.5 | $10.00 | |
| 2 | Claude Opus 5Anthropic | 1533 | 1,597 | ±15.1 | 81.8 | $25.00 | |
| 3 | Gemini 3.8 FlashGoogle | 1532 | 2,094 | ±13.4 | 70.5 | $4.125 | |
| 4 | Claude Fable 5.1Anthropic | 1531 | 746 | ±21.1 | 89.1 | $50.00 | |
| 5 | Claude Opus 5.5Anthropic | 1518 | 491 | ±26.2 | 92.2 | $20.00 | |
| 6 | Claude Opus 4.6Anthropic | 1518 | 4,248 | ±9.5 | 66.9 | $25.00 | |
| 7 | Claude Fable 5Anthropic | 1516 | 2,161 | ±12.9 | 82.1 | $50.00 | |
| 8 | Gemini 3.7 FlashGoogle | 1511 | 1,501 | ±15.5 | 69.9 | $3.75 | |
| 9 | Gemini 3.6 FlashGoogle | 1508 | 1,760 | ±14.1 | 48.3 | $3.75 | |
| 10 | Gemini 3.5 FlashGoogle | 1504 | 2,388 | ±12.4 | 42.8 | $9.00 | |
| 11 | GLM 5.3 FlashZ.ai | 1503 | 1,820 | ±14.2 | 71.0 | $0.50 | |
| 12 | Qwen3.8 Max (0902)Qwen | 1501 | 1,481 | ±15.5 | 63.5 | $6.00 | |
| 13 | MiMo-V2.6-ProXiaomi | 1500 | 215 | ±39.8 | 72.6 | $0.87 | |
| 14 | Claude Opus 4.7Anthropic | 1498 | 3,293 | ±10.7 | 51.1 | $25.00 | |
| 15 | Muse Spark 1.3Meta | 1497 | 948 | ±18.9 | 79.7 | $4.25 | |
| 16 | Claude Sonnet 5.5Anthropic | 1497 | 416 | ±28.4 | 88.9 | $10.00 | |
| 17 | GLM 5.3Z.ai | 1495 | 1,023 | ±18.0 | 73.1 | $1.40 | |
| 18 | Kimi K3Moonshot AI | 1494 | 1,535 | ±15.1 | 61.8 | $11.936 | |
| 19 | GPT-6.1 SolOpenAI | 1489 | 390 | ±28.4 | 88.7 | $10.00 | |
| 20 | GPT-5.4OpenAI | 1488 | 3,377 | ±10.6 | 37.7 | $15.00 | |
| 21 | GPT-5.5OpenAI | 1487 | 3,518 | ±10.4 | 67.8 | $30.00 | |
| 22 | Claude Opus 4.8Anthropic | 1486 | 2,974 | ±11.2 | 66.0 | $25.00 | |
| 23 | Muse WonderMeta | 1486 | 277 | ±33.3 | 64.5 | $20.00 | |
| 24 | Gemini 3.1 Pro PreviewGoogle | 1486 | 6,734 | ±7.9 | 35.3 | $12.00 | |
| 25 | Muse Spark 1.2Meta | 1484 | 484 | ±26.1 | 65.1 | $4.25 | |
| 26 | GLM 5.2Z.ai | 1483 | 2,303 | ±12.4 | 57.7 | $7.00 | |
| 27 | Muse Spark 1.1Meta | 1482 | 2,076 | ±13.3 | 48.9 | $4.25 | |
| 28 | MiMo-V2.5-ProXiaomi | 1481 | 3,709 | ±10.1 | 40.0 | $0.609 | |
| 29 | T InklingThinking Machines | 1479 | 1,951 | ±13.4 | 49.7 | $4.05 | |
| 30 | DeepSeek V4.1 FlashDeepSeek | 1478 | 704 | ±21.6 | 77.2 | $0.525 | |
| 31 | MiMo-V2.6-FlashXiaomi | 1476 | 676 | ±21.9 | 61.9 | $0.28 | |
| 32 | Kimi K2.6Moonshot AI | 1474 | 2,072 | ±13.1 | 44.2 | $3.41 | |
| 33 | Hy3Tencent | 1473 | 689 | ±21.8 | 43.9 | $0.528 | |
| 34 | GPT-5.6 SolOpenAI | 1473 | 1,981 | ±13.4 | 76.7 | $20.00 | |
| 35 | GLM 5.1Z.ai | 1473 | 3,182 | ±10.9 | 44.0 | $4.40 | |
| 36 | Kimi K2.5Moonshot AI | 1470 | 4,174 | ±9.3 | 58.5 | $2.50 | |
| 37 | Gemma 4 26B A4BGoogle | 1470 | 390 | ±27.5 | 22.8 | $0.255 | |
| 38 | Mistral Large 4Mistral AI | 1469 | 488 | ±26.0 | 55.3 | $4.18 | |
| 39 | Claude Sonnet 5Anthropic | 1467 | 2,325 | ±12.4 | 63.5 | $10.00 | |
| 40 | Qwen3.7 PlusQwen | 1467 | 1,917 | ±13.7 | Not benchmarked | $1.28 | |
| 41 | GPT-6 AstraOpenAI | 1465 | 687 | ±22.2 | 88.7 | $50.00 | |
| 42 | GPT-5.6 TerraOpenAI | 1465 | 2,056 | ±13.1 | 69.0 | $12.00 | |
| 43 | Claude Opus 4.5Anthropic | 1463 | 2,346 | ±12.1 | 66.8 | $25.00 | |
| 44 | Gemma 4 31BGoogle | 1463 | 428 | ±26.6 | 38.3 | $0.40 | |
| 45 | Claude Sonnet 4.6Anthropic | 1462 | 3,796 | ±10.0 | 57.8 | $15.00 | |
| 46 | Grok 4.5xAI | 1460 | 1,765 | ±14.2 | 75.5 | $6.00 | |
| 47 | GPT-5.6 LunaOpenAI | 1459 | 1,926 | ±13.7 | 61.5 | $1.20 | |
| 48 | T Inkling SmallThinking Machines | 1459 | 1,556 | ±15.1 | 29.9 | $1.20 | |
| 49 | R Beam 1.0 501B A23BReflection | 1459 | 233 | ±36.5 | 33.8 | $1.00 | |
| 50 | Step 5 PreviewStepFun | 1458 | 443 | ±26.6 | 71.4 | $2.70 | |
| 51 | Nemotron 3 UltraNVIDIA | 1457 | 580 | ±24.1 | Not benchmarked | $2.40 | |
| 52 | Qwen3.8 27BQwen | 1455 | 1,250 | ±16.7 | 73.6 | $1.35 | |
| 53 | Qwen3.5 397B A17BQwen | 1453 | 5,233 | ±8.6 | 28.6 | $2.34 | |
| 54 | Gemini 2.5 ProGoogle | 1450 | 7,818 | ±6.9 | 37.0 | $10.00 | |
| 55 | Qwen3.6 PlusQwen | 1449 | 2,685 | ±11.5 | 37.4 | $1.95 | |
| 56 | Claude Sonnet 4.5Anthropic | 1449 | 5,075 | ±8.4 | 59.1 | $15.00 | |
| 57 | GPT-5.1OpenAI | 1446 | 2,580 | ±11.6 | 53.3 | $10.00 | |
| 58 | DeepSeek V4 Pro 0423DeepSeek | 1443 | 3,068 | ±11.1 | 35.3 | $1.9105 | |
| 59 | GLM 5V TurboZ.ai | 1442 | 485 | ±25.1 | 28.8 | $4.00 | |
| 60 | DeepSeek V4 Pro 0813DeepSeek | 1441 | 407 | ±27.8 | 54.6 | $1.98 | |
| 61 | GLM 5Z.ai | 1440 | 1,694 | ±14.1 | 59.7 | $3.20 | |
| 62 | Qwen3 Next 80B A3B InstructQwen | 1440 | 1,204 | ±16.7 | Not benchmarked | $1.10 | |
| 63 | GPT-5.2OpenAI | 1440 | 3,125 | ±10.7 | 58.9 | $14.00 | |
| 64 | MiMo-V2.5Xiaomi | 1436 | 2,461 | ±12.1 | 34.0 | $0.238 | |
| 65 | DeepSeek V3.2DeepSeek | 1435 | 3,140 | ±10.6 | 54.4 | $0.42 | |
| 66 | Gemini 3.5 Flash LiteGoogle | 1434 | 1,622 | ±14.7 | 33.6 | $2.50 | |
| 67 | GLM 4.6Z.ai | 1432 | 2,154 | ±12.5 | 52.2 | $2.20 | |
| 68 | Qwen3.5-122B-A10BQwen | 1431 | 1,842 | ±13.7 | 22.9 | $2.08 | |
| 69 | Mistral Medium 3.5Mistral AI | 1431 | 565 | ±24.3 | 9.2 | $7.50 | |
| 70 | Claude Opus 4.1Anthropic | 1431 | 3,058 | ±10.5 | 27.2 | $75.00 | |
| 71 | MiniMax M3MiniMax | 1429 | 3,307 | ±10.7 | 57.8 | $1.20 | |
| 72 | Qwen3.5-27BQwen | 1429 | 1,745 | ±14.0 | 22.6 | $2.16 | |
| 73 | Gemini 3.1 Flash Lite PreviewGoogle | 1428 | 3,619 | ±10.1 | 8.1 | $1.50 | |
| 74 | GLM 4.5Z.ai | 1427 | 1,401 | ±15.6 | 64.2 | $2.20 | |
| 75 | o3OpenAI | 1426 | 3,701 | ±9.9 | 70.7 | $8.00 | |
| 76 | DeepSeek V4 Flash 0423DeepSeek | 1426 | 2,699 | ±11.7 | 54.6 | $0.182 | |
| 77 | Qwen3 VL 235B A22B InstructQwen | 1425 | 699 | ±22.7 | Not benchmarked | $1.90 | |
| 78 | Hy3 previewTencent | 1424 | 430 | ±26.7 | 22.4 | $0.60 | |
| 79 | GLM 4.7Z.ai | 1424 | 738 | ±20.4 | 33.6 | $1.98 | |
| 80 | DeepSeek V3.2 ExpDeepSeek | 1422 | 477 | ±26.5 | 55.0 | $0.41 | |
| 81 | Solar Pro 4Upstage | 1421 | 1,181 | ±17.2 | 24.4 | $0.60 | |
| 82 | MiniMax M2.7MiniMax | 1421 | 4,372 | ±9.5 | 28.4 | $1.20 | |
| 83 | GPT-5.4 MiniOpenAI | 1418 | 3,369 | ±10.7 | 28.2 | $4.50 | |
| 84 | Grok 4.6xAI | 1417 | 1,521 | ±15.3 | 76.4 | $6.00 | |
| 85 | Grok 4.7xAI | 1414 | 693 | ±21.9 | 75.0 | $6.00 | |
| 86 | Qwen3 235B A22B Thinking 2507Qwen | 1412 | 479 | ±24.7 | Not benchmarked | $3.50 | |
| 87 | Gemini 2.5 FlashGoogle | 1411 | 8,004 | ±6.6 | 28.7 | $2.50 | |
| 88 | GPT-6 LunaOpenAI | 1409 | 3,477 | ±11.0 | 71.4 | $0.50 | |
| 89 | Mistral Medium 3.1Mistral AI | 1408 | 5,994 | ±7.7 | Not benchmarked | $2.00 | |
| 90 | Step 3.5 FlashStepFun | 1408 | 3,424 | ±10.2 | Not benchmarked | $0.30 | |
| 91 | GPT-6 SolOpenAI | 1407 | 767 | ±21.3 | 82.8 | $10.00 | |
| 92 | Qwen3.5-FlashQwen | 1407 | 3,388 | ±10.5 | 6.2 | $0.26 | |
| 93 | GPT-5OpenAI | 1407 | 1,780 | ±13.7 | 72.2 | $10.00 | |
| 94 | GPT-5.4 NanoOpenAI | 1405 | 3,261 | ±10.8 | Not benchmarked | $1.25 | |
| 95 | Qwen3.5-35B-A3BQwen | 1404 | 1,850 | ±13.7 | 7.7 | $1.00 | |
| 96 | DeepSeek V3.1 TerminusDeepSeek | 1402 | 216 | ±38.5 | Not benchmarked | $1.00 | |
| 97 | R1 0528DeepSeek | 1400 | 861 | ±19.9 | 71.4 | $2.15 | |
| 98 | Qwen3 Next 80B A3B ThinkingQwen | 1399 | 822 | ±19.7 | Not benchmarked | $1.20 | |
| 99 | Qwen3 32BQwen | 1399 | 316 | ±30.0 | 40.0 | $0.28 | |
| 100 | Claude Haiku 4.5Anthropic | 1398 | 8,434 | ±7.0 | 49.8 | $5.00 | |
| 101 | o3 MiniOpenAI | 1396 | 1,909 | ±13.2 | 52.1 | $4.40 | |
| 102 | GLM 4.5 AirZ.ai | 1396 | 1,528 | ±14.8 | Not benchmarked | $0.85 | |
| 103 | Qwen3 30B A3B Instruct 2507Qwen | 1394 | 1,394 | ±15.1 | Not benchmarked | $0.30 | |
| 104 | o4 MiniOpenAI | 1389 | 2,916 | ±10.7 | 68.6 | $4.40 | |
| 105 | gpt-oss-120bOpenAI | 1389 | 1,781 | ±13.8 | 34.5 | $0.60 | |
| 106 | o1OpenAI | 1388 | 2,986 | ±10.8 | 61.7 | $60.00 | |
| 107 | Grok 4.3xAI | 1387 | 3,418 | ±10.7 | 22.5 | $2.50 | |
| 108 | Nemotron 3 SuperNVIDIA | 1380 | 547 | ±24.4 | Not benchmarked | $2.40 | |
| 109 | MiniMax M2.5MiniMax | 1378 | 2,551 | ±11.8 | 59.4 | $0.95 | |
| 110 | GPT-5 MiniOpenAI | 1378 | 1,458 | ±15.4 | 59.8 | $2.00 | |
| 111 | Nemotron 3 Nano 30B A3BNVIDIA | 1371 | 1,014 | ±18.0 | Not benchmarked | $0.24 | |
| 112 | GPT-4.1OpenAI | 1369 | 3,227 | ±10.2 | 46.5 | $8.00 | |
| 113 | Qwen3 Coder 480B A35BQwen | 1365 | 1,610 | ±14.8 | 50.0 | $1.00 | |
| 114 | Trinity Large ThinkingArcee AI | 1365 | 1,723 | ±14.8 | 5.4 | $0.80 | |
| 115 | MiniMax M1MiniMax | 1361 | 1,771 | ±13.5 | Not benchmarked | $2.20 | |
| 116 | Nova 2 LiteAmazon | 1359 | 861 | ±19.6 | Not benchmarked | $2.50 | |
| 117 | GLM 4.7 FlashZ.ai | 1355 | 756 | ±20.6 | Not benchmarked | $0.40 | |
| 118 | Qwen3 30B A3BQwen | 1355 | 1,706 | ±13.5 | Not benchmarked | $0.50 | |
| 119 | GLM 4.5VZ.ai | 1354 | 269 | ±34.2 | Not benchmarked | $1.80 | |
| 120 | Mistral Medium 3Mistral AI | 1353 | 2,233 | ±12.0 | Not benchmarked | $2.00 | |
| 121 | MiniMax M2MiniMax | 1351 | 320 | ±32.8 | 45.3 | $1.02 | |
| 122 | GPT-4.1 MiniOpenAI | 1343 | 2,674 | ±11.1 | 28.5 | $1.60 | |
| 123 | Qwen-PlusQwen | 1326 | 732 | ±19.3 | Not benchmarked | $0.78 | |
| 124 | GPT-5 NanoOpenAI | 1317 | 494 | ±26.5 | 34.8 | $0.40 | |
| 125 | gpt-oss-20bOpenAI | 1317 | 678 | ±22.4 | Not benchmarked | $0.09 | |
| 126 | Gemma 3 27BGoogle | 1312 | 3,587 | ±9.5 | 4.9 | $0.35 | |
| 127 | Gemma 3 12BGoogle | 1307 | 389 | ±26.9 | Not benchmarked | $0.15 | |
| 128 | Command ACohere | 1300 | 3,993 | ±9.2 | Not benchmarked | $10.00 | |
| 129 | Llama 4 MaverickMeta Llama | 1299 | 2,825 | ±10.8 | 18.1 | $0.6525 | |
| 130 | Llama 4 ScoutMeta Llama | 1287 | 1,942 | ±13.2 | 9.1 | $0.30 | |
| 131 | GPT-4o (2024-08-06)OpenAI | 1285 | 6,826 | ±8.1 | 23.1 | $10.00 | |
| 132 | GPT-4o (2024-05-13)OpenAI | 1284 | 15,103 | ±6.9 | Not benchmarked | $15.00 | |
| 133 | Qwen2.5 72B InstructQwen | 1283 | 5,415 | ±8.3 | Not benchmarked | $0.40 | |
| 134 | GPT-4.1 NanoOpenAI | 1274 | 582 | ±23.0 | 8.9 | $0.40 | |
| 135 | GPT-4 TurboOpenAI | 1272 | 13,217 | ±7.6 | Not benchmarked | $30.00 | |
| 136 | GPT-4o-mini (2024-07-18)OpenAI | 1267 | 9,322 | ±7.0 | 3.6 | $0.60 | |
| 137 | Llama 3.3 70B InstructMeta Llama | 1267 | 5,779 | ±7.7 | Not benchmarked | $0.32 | |
| 138 | Mistral Small 3.1 24BMistral AI | 1261 | 2,105 | ±13.1 | Not benchmarked | $0.555 | |
| 139 | Mistral LargeMistral AI | 1261 | 3,574 | ±9.3 | Not benchmarked | $6.00 | |
| 140 | Llama 3.1 70B InstructMeta Llama | 1252 | 7,677 | ±7.6 | Not benchmarked | $0.40 | |
| 141 | Nova Pro 1.0Amazon | 1252 | 2,978 | ±10.0 | Not benchmarked | $3.20 | |
| 142 | Qwen2.5 Coder 32B InstructQwen | 1251 | 725 | ±19.2 | 30.5 | $1.00 | |
| 143 | Phi 4Microsoft | 1246 | 2,764 | ±10.5 | Not benchmarked | $0.14 | |
| 144 | Mistral Small 3Mistral AI | 1240 | 1,683 | ±13.4 | Not benchmarked | $0.08 | |
| 145 | Gemma 3 4BGoogle | 1239 | 423 | ±27.9 | Not benchmarked | $0.05 | |
| 146 | GPT-4OpenAI | 1230 | 7,052 | ±9.7 | 22.4 | $60.00 | |
| 147 | Nova Lite 1.0Amazon | 1227 | 2,511 | ±10.9 | Not benchmarked | $0.24 | |
| 148 | Gemma 2 27BGoogle | 1212 | 10,170 | ±6.9 | Not benchmarked | $0.65 | |
| 149 | Nova Micro 1.0Amazon | 1206 | 2,455 | ±10.9 | Not benchmarked | $0.14 | |
| 150 | Command R (08-2024)Cohere | 1188 | 1,467 | ±13.8 | Not benchmarked | $0.60 | |
| 151 | Llama 3.1 8B InstructMeta Llama | 1179 | 7,135 | ±8.0 | Not benchmarked | $0.08 | |
| 152 | Llama 3.2 3B InstructMeta Llama | 1126 | 1,136 | ±15.7 | Not benchmarked | $0.335 | |
| 153 | Llama 3.2 1B InstructMeta Llama | 1086 | 1,162 | ±16.0 | 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.