LMArena Text: science leaderboard
- Gemini 4 Argon leads
- 1548
- models measured
- 156
- lead over #2
- +15 Elo
- median
- 1430
Leader on LMArena Text: science · read October 22
On October 22, 2026, 168 of our catalog's models had a LMArena Text: science result, and Google Gemini 4 Argon led with 1548, ahead of Anthropic Claude Fable 5.1 at 1532.
Prompts from the life, physical and social sciences. 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
LMArena Text: science is shown as published and does not feed any of our indexes. Read the methodology
| # | Model | Result | Votes | Interval (±) | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|---|---|
| 1 | Gemini 4 ArgonGoogle | 1548 | 818 | ±21.1 | 81.0 | $10.00 | |
| 2 | Claude Fable 5.1Anthropic | 1532 | 2,595 | ±12.0 | 84.2 | $50.00 | |
| 3 | Claude Opus 5Anthropic | 1531 | 5,659 | ±8.5 | 83.0 | $25.00 | |
| 4 | Claude Opus 5.5Anthropic | 1524 | 1,644 | ±15.2 | 88.8 | $20.00 | |
| 5 | Gemini 3.8 FlashGoogle | 1523 | 6,873 | ±8.0 | 61.6 | $3.75 | |
| 6 | Claude Opus 4.6Anthropic | 1519 | 13,772 | ±6.0 | 57.2 | $25.00 | |
| 7 | Qwen3.8 Max (0902)Qwen | 1512 | 4,599 | ±9.2 | 62.4 | $6.00 | |
| 8 | Gemini 3.7 FlashGoogle | 1511 | 4,871 | ±9.0 | 58.0 | $3.75 | |
| 9 | Kimi K3Moonshot AI | 1508 | 5,364 | ±8.5 | 62.3 | $15.00 | |
| 10 | Claude Opus 4.7Anthropic | 1508 | 11,056 | ±6.7 | 67.5 | $25.00 | |
| 11 | Muse Spark 1.3Meta | 1507 | 3,147 | ±11.0 | 57.4 | $4.25 | |
| 12 | Claude Fable 5Anthropic | 1506 | 7,086 | ±7.7 | 79.3 | $50.00 | |
| 13 | Muse WonderMeta | 1506 | 824 | ±20.6 | 71.9 | $20.00 | |
| 14 | MiMo-V2.6-ProXiaomi | 1506 | 601 | ±23.2 | 65.4 | $0.87 | |
| 15 | Muse Spark 1.2Meta | 1500 | 1,537 | ±15.2 | 55.7 | $4.25 | |
| 16 | GLM 5.3Z.ai | 1498 | 4,055 | ±9.7 | 70.6 | $4.40 | |
| 17 | Gemini 3.1 Pro PreviewGoogle | 1497 | 21,617 | ±5.1 | 32.3 | $12.00 | |
| 18 | Gemini 3.5 FlashGoogle | 1495 | 7,779 | ±7.4 | 39.8 | $9.00 | |
| 19 | GLM 5.2Z.ai | 1489 | 8,550 | ±7.1 | 61.1 | $4.40 | |
| 20 | Gemini 3.6 FlashGoogle | 1489 | 5,809 | ±8.2 | 62.5 | $3.75 | |
| 21 | GPT-5.5OpenAI | 1485 | 11,611 | ±6.5 | 71.3 | $30.00 | |
| 22 | Muse Spark 1.1Meta | 1484 | 7,539 | ±7.5 | 70.3 | $4.25 | |
| 23 | GLM 5.3 FlashZ.ai | 1484 | 5,949 | ±8.3 | 57.7 | $0.50 | |
| 24 | Claude Sonnet 5.5Anthropic | 1483 | 1,414 | ±16.3 | 87.3 | $10.00 | |
| 25 | MiMo-V2.5-ProXiaomi | 1482 | 12,756 | ±6.1 | 31.6 | $0.9152 | |
| 26 | Gemini 2.5 ProGoogle | 1480 | 20,872 | ±4.6 | 19.1 | $10.00 | |
| 27 | GPT-5.4OpenAI | 1479 | 10,646 | ±6.7 | 64.4 | $15.00 | |
| 28 | GLM 5.1Z.ai | 1478 | 10,616 | ±6.4 | 49.2 | $4.40 | |
| 29 | Claude Opus 4.8Anthropic | 1477 | 10,443 | ±6.7 | 65.0 | $25.00 | |
| 30 | Claude Sonnet 4.6Anthropic | 1474 | 11,659 | ±6.4 | 57.3 | $15.00 | |
| 31 | DeepSeek V4 Pro 0423DeepSeek | 1470 | 9,159 | ±7.1 | 38.7 | $2.976 | |
| 32 | DeepSeek V4.1 FlashDeepSeek | 1469 | 2,203 | ±12.8 | 62.5 | $0.84 | |
| 33 | Kimi K2.6Moonshot AI | 1468 | 6,716 | ±7.8 | 52.1 | $4.00 | |
| 34 | Qwen3.8 27BQwen | 1467 | 4,001 | ±9.8 | 74.2 | $2.55 | |
| 35 | Nemotron 3 UltraNVIDIA | 1465 | 1,893 | ±13.7 | 30.0 | $2.40 | |
| 36 | Grok 4.5xAI | 1464 | 6,533 | ±7.8 | 58.6 | $6.00 | |
| 37 | Qwen3.7 PlusQwen | 1464 | 6,753 | ±7.7 | 33.2 | $1.28 | |
| 38 | MiMo-V2.6-FlashXiaomi | 1462 | 1,967 | ±13.5 | 54.7 | $0.28 | |
| 39 | GLM 5Z.ai | 1460 | 4,720 | ±8.7 | 41.7 | $3.20 | |
| 40 | Claude Sonnet 5Anthropic | 1460 | 8,423 | ±7.2 | 57.7 | $10.00 | |
| 41 | GPT-5.1OpenAI | 1459 | 6,884 | ±7.6 | 40.4 | $10.00 | |
| 42 | Kimi K2.5Moonshot AI | 1459 | 11,807 | ±6.0 | 38.6 | $2.85 | |
| 43 | GLM 4.7Z.ai | 1457 | 2,139 | ±12.5 | 55.1 | $2.20 | |
| 44 | Hy3Tencent | 1457 | 2,014 | ±13.2 | Not benchmarked | $0.528 | |
| 45 | Gemma 4 31BGoogle | 1457 | 931 | ±18.8 | 39.0 | $0.40 | |
| 46 | Qwen3.5 397B A17BQwen | 1457 | 15,562 | ±5.5 | Not benchmarked | $2.34 | |
| 47 | DeepSeek V4 Pro 0813DeepSeek | 1457 | 1,527 | ±15.1 | 56.1 | $1.98 | |
| 48 | GLM 4.6Z.ai | 1456 | 5,928 | ±7.9 | 8.8 | $2.20 | |
| 49 | T InklingThinking Machines | 1455 | 6,975 | ±7.7 | 31.5 | $4.05 | |
| 50 | GPT-5.6 SolOpenAI | 1455 | 6,789 | ±7.8 | 73.5 | $20.00 | |
| 51 | Step 5 PreviewStepFun | 1454 | 1,416 | ±15.6 | 61.1 | $2.70 | |
| 52 | Claude Opus 4.5Anthropic | 1454 | 11,580 | ±6.1 | 44.0 | $25.00 | |
| 53 | GPT-5.6 TerraOpenAI | 1452 | 7,387 | ±7.6 | 63.9 | $12.00 | |
| 54 | MiniMax M3MiniMax | 1452 | 10,438 | ±6.6 | 39.2 | $1.20 | |
| 55 | DeepSeek V3.1 TerminusDeepSeek | 1451 | 571 | ±24.2 | Not benchmarked | $1.0142 | |
| 56 | GPT-6 AstraOpenAI | 1451 | 2,266 | ±12.8 | 85.1 | $50.00 | |
| 57 | Gemini 3.5 Flash LiteGoogle | 1451 | 5,852 | ±8.3 | 49.0 | $2.50 | |
| 58 | Mistral Large 4Mistral AI | 1448 | 1,541 | ±15.3 | 56.4 | $4.18 | |
| 59 | GLM 5V TurboZ.ai | 1448 | 1,646 | ±14.5 | Not benchmarked | $4.00 | |
| 60 | Claude Sonnet 4.5Anthropic | 1448 | 13,618 | ±5.6 | 39.5 | $15.00 | |
| 61 | DeepSeek V4 Flash 0423DeepSeek | 1446 | 8,396 | ±7.3 | Not benchmarked | $0.28 | |
| 62 | Gemma 4 26B A4BGoogle | 1446 | 926 | ±18.5 | Not benchmarked | $0.60 | |
| 63 | DeepSeek V3.2DeepSeek | 1446 | 7,628 | ±7.2 | 30.0 | $0.88 | |
| 64 | Qwen3 VL 235B A22B InstructQwen | 1444 | 1,734 | ±15.2 | Not benchmarked | $1.90 | |
| 65 | Mistral Medium 3.1Mistral AI | 1444 | 15,297 | ±5.2 | Not benchmarked | $2.00 | |
| 66 | GLM 4.5Z.ai | 1444 | 3,861 | ±9.7 | Not benchmarked | $2.20 | |
| 67 | GPT-6.1 SolOpenAI | 1444 | 1,282 | ±16.8 | 82.4 | $10.00 | |
| 68 | DeepSeek V3.2 ExpDeepSeek | 1443 | 1,424 | ±15.0 | Not benchmarked | $0.41 | |
| 69 | MiMo-V2.5Xiaomi | 1440 | 7,709 | ±7.6 | 50.0 | $0.336 | |
| 70 | Qwen3.6 PlusQwen | 1439 | 7,641 | ±7.5 | 44.1 | $1.95 | |
| 71 | Qwen3.5-27BQwen | 1437 | 4,780 | ±8.7 | Not benchmarked | $1.56 | |
| 72 | Grok 4.6xAI | 1437 | 5,140 | ±8.8 | 65.7 | $6.00 | |
| 73 | Qwen3 Next 80B A3B InstructQwen | 1434 | 3,530 | ±10.0 | Not benchmarked | $1.10 | |
| 74 | GPT-5.6 LunaOpenAI | 1433 | 6,718 | ±7.9 | 58.0 | $1.20 | |
| 75 | Qwen3 235B A22B Thinking 2507Qwen | 1433 | 1,494 | ±15.2 | Not benchmarked | $3.50 | |
| 76 | Gemini 2.5 FlashGoogle | 1433 | 20,586 | ±4.6 | Not benchmarked | $2.50 | |
| 77 | Qwen3.5-122B-A10BQwen | 1432 | 4,915 | ±8.6 | Not benchmarked | $2.08 | |
| 78 | R1 0528DeepSeek | 1430 | 3,387 | ±10.6 | Not benchmarked | $2.18 | |
| 79 | R Beam 1.0 501B A23BReflection | 1430 | 686 | ±23.0 | Not benchmarked | $1.10 | |
| 80 | o3OpenAI | 1429 | 9,923 | ±6.7 | Not benchmarked | $8.00 | |
| 81 | Mistral Medium 3.5Mistral AI | 1428 | 1,938 | ±13.6 | 25.5 | $7.50 | |
| 82 | T Inkling SmallThinking Machines | 1428 | 5,673 | ±8.4 | 27.3 | $1.20 | |
| 83 | Gemini 3.1 Flash Lite PreviewGoogle | 1425 | 10,505 | ±6.5 | 22.7 | $1.50 | |
| 84 | GPT-5.2OpenAI | 1425 | 8,273 | ±7.0 | 55.9 | $14.00 | |
| 85 | Claude Opus 4.1Anthropic | 1422 | 7,939 | ±7.0 | Not benchmarked | $75.00 | |
| 86 | Grok 4.7xAI | 1419 | 2,283 | ±12.7 | 66.7 | $6.00 | |
| 87 | GPT-5.4 MiniOpenAI | 1417 | 10,533 | ±6.8 | 35.8 | $4.50 | |
| 88 | MiniMax M2.7MiniMax | 1416 | 13,861 | ±5.9 | 36.2 | $1.20 | |
| 89 | GPT-5OpenAI | 1416 | 5,011 | ±8.5 | 38.8 | $10.00 | |
| 90 | Hy3 previewTencent | 1416 | 1,241 | ±17.0 | Not benchmarked | $0.60 | |
| 91 | Qwen3.5-FlashQwen | 1415 | 9,843 | ±6.7 | 41.1 | $0.26 | |
| 92 | Step 3.5 FlashStepFun | 1415 | 9,816 | ±6.5 | Not benchmarked | $0.30 | |
| 93 | Qwen3.5-35B-A3BQwen | 1413 | 5,205 | ±8.4 | Not benchmarked | $1.30 | |
| 94 | GLM 4.5 AirZ.ai | 1407 | 4,923 | ±8.5 | Not benchmarked | $1.10 | |
| 95 | Claude Haiku 4.5Anthropic | 1406 | 24,635 | ±4.6 | 35.6 | $5.00 | |
| 96 | Solar Pro 4Upstage | 1404 | 3,855 | ±10.1 | Not benchmarked | $0.60 | |
| 97 | Grok 4.3xAI | 1399 | 11,817 | ±6.5 | 37.4 | $2.50 | |
| 98 | Qwen3 30B A3B Instruct 2507Qwen | 1393 | 3,677 | ±9.8 | Not benchmarked | $0.30 | |
| 99 | GLM 4.6VZ.ai | 1389 | 480 | ±26.4 | Not benchmarked | $0.90 | |
| 100 | GPT-6 SolOpenAI | 1389 | 2,325 | ±12.9 | 76.0 | $10.00 | |
| 101 | Nemotron 3 SuperNVIDIA | 1387 | 1,221 | ±16.2 | Not benchmarked | $0.45 | |
| 102 | GPT-6 LunaOpenAI | 1386 | 10,709 | ±7.0 | 57.3 | $0.50 | |
| 103 | Qwen3 Next 80B A3B ThinkingQwen | 1384 | 2,104 | ±12.8 | Not benchmarked | $1.20 | |
| 104 | Gemma 3 27BGoogle | 1383 | 7,868 | ±7.1 | Not benchmarked | $0.30 | |
| 105 | Nova 2 LiteAmazon | 1382 | 2,131 | ±12.9 | Not benchmarked | $2.50 | |
| 106 | GPT-4.1OpenAI | 1382 | 8,489 | ±6.9 | Not benchmarked | $8.00 | |
| 107 | gpt-oss-120bOpenAI | 1382 | 4,829 | ±8.8 | Not benchmarked | $0.50 | |
| 108 | Mistral Medium 3Mistral AI | 1381 | 5,634 | ±8.3 | Not benchmarked | $2.00 | |
| 109 | R1RetiredDeepSeek | 1375 | 3,309 | ±10.3 | Not benchmarked | $2.50 | |
| 110 | GPT-5.4 NanoOpenAI | 1374 | 10,519 | ±6.9 | 44.8 | $1.25 | |
| 111 | Qwen3 32BQwen | 1372 | 798 | ±20.2 | Not benchmarked | $0.40 | |
| 112 | GPT-5 MiniOpenAI | 1371 | 4,211 | ±9.3 | 34.2 | $2.00 | |
| 113 | MiniMax M2.5MiniMax | 1369 | 7,140 | ±7.5 | 37.2 | $1.20 | |
| 114 | o1OpenAI | 1369 | 4,820 | ±8.8 | Not benchmarked | $60.00 | |
| 115 | Gemma 3 12BGoogle | 1366 | 760 | ±20.5 | Not benchmarked | $0.1625 | |
| 116 | Claude Opus 4RetiredAnthropic | 1363 | 6,982 | ±7.8 | Not benchmarked | $75.00 | |
| 117 | Gemini 2.0 FlashRetiredGoogle | 1360 | 7,782 | ±7.1 | Not benchmarked | $0.40 | |
| 118 | GLM 4.7 FlashZ.ai | 1358 | 1,878 | ±13.3 | Not benchmarked | $0.40 | |
| 119 | Trinity Large ThinkingArcee AI | 1355 | 5,100 | ±9.0 | Not benchmarked | $0.80 | |
| 120 | o4 MiniOpenAI | 1354 | 7,538 | ±7.3 | Not benchmarked | $4.40 | |
| 121 | MiniMax M1MiniMax | 1354 | 5,799 | ±8.0 | Not benchmarked | $2.20 | |
| 122 | Nemotron 3 Nano 30B A3BNVIDIA | 1353 | 2,551 | ±11.9 | Not benchmarked | $0.22 | |
| 123 | Qwen-PlusQwen | 1352 | 968 | ±18.5 | Not benchmarked | $0.78 | |
| 124 | MiniMax M2MiniMax | 1348 | 1,124 | ±17.8 | Not benchmarked | $1.02 | |
| 125 | Qwen3 Coder 480B A35BQwen | 1347 | 4,028 | ±9.8 | Not benchmarked | $1.50 | |
| 126 | Mercury 2Inception | 1343 | 492 | ±26.0 | Not benchmarked | $0.75 | |
| 127 | o3 MiniOpenAI | 1339 | 3,382 | ±10.4 | Not benchmarked | $4.40 | |
| 128 | Command ACohere | 1339 | 9,387 | ±6.5 | Not benchmarked | $10.00 | |
| 129 | Granite 4.2 8BIBM Granite | 1336 | 523 | ±26.6 | Not benchmarked | $0.20 | |
| 130 | GLM 4.5VZ.ai | 1330 | 735 | ±21.4 | Not benchmarked | $1.80 | |
| 131 | GPT-4.1 MiniOpenAI | 1328 | 6,493 | ±7.9 | Not benchmarked | $1.60 | |
| 132 | Qwen3 30B A3BQwen | 1321 | 4,583 | ±9.1 | Not benchmarked | $0.50 | |
| 133 | Gemini 1.5 ProRetiredGoogle | 1321 | 9,790 | ±6.9 | Not benchmarked | $5.00 | |
| 134 | GPT-5 NanoOpenAI | 1318 | 1,309 | ±16.3 | Not benchmarked | $0.40 | |
| 135 | o1-miniRetiredOpenAI | 1316 | 8,979 | ±7.3 | Not benchmarked | $4.40 | |
| 136 | Gemma 3 4BGoogle | 1306 | 780 | ±20.6 | Not benchmarked | $0.075 | |
| 137 | Gemini 1.5 FlashRetiredGoogle | 1299 | 6,083 | ±8.3 | Not benchmarked | $0.30 | |
| 138 | GPT-4.1 NanoOpenAI | 1297 | 1,162 | ±16.3 | Not benchmarked | $0.40 | |
| 139 | gpt-oss-20bOpenAI | 1296 | 1,668 | ±14.9 | Not benchmarked | $0.15 | |
| 140 | GPT-4o (2024-05-13)OpenAI | 1295 | 19,017 | ±6.7 | Not benchmarked | $15.00 | |
| 141 | Llama 4 MaverickMeta Llama | 1288 | 6,714 | ±7.9 | Not benchmarked | $1.00 | |
| 142 | GPT-4o-mini (2024-07-18)OpenAI | 1286 | 11,641 | ±6.6 | Not benchmarked | $0.60 | |
| 143 | Claude 3.7 SonnetRetiredAnthropic | 1285 | 7,445 | ±7.7 | Not benchmarked | $15.00 | |
| 144 | Llama 3.3 70B InstructMeta Llama | 1284 | 9,353 | ±6.5 | Not benchmarked | $0.72 | |
| 145 | Llama 4 ScoutMeta Llama | 1278 | 4,978 | ±9.0 | Not benchmarked | $0.50 | |
| 146 | Llama 3.1 70B InstructMeta Llama | 1276 | 9,381 | ±7.4 | Not benchmarked | $0.56 | |
| 147 | Qwen2.5 72B InstructQwen | 1275 | 6,932 | ±7.9 | Not benchmarked | $0.40 | |
| 148 | GPT-4o (2024-08-06)OpenAI | 1274 | 7,560 | ±8.0 | Not benchmarked | $10.00 | |
| 149 | Claude 3.5 SonnetRetiredAnthropic | 1270 | 15,553 | ±5.7 | Not benchmarked | $15.00 | |
| 150 | Mistral Small 3.1 24BMistral AI | 1265 | 5,222 | ±8.8 | Not benchmarked | $0.555 | |
| 151 | Mistral LargeMistral AI | 1260 | 4,869 | ±8.7 | Not benchmarked | $6.00 | |
| 152 | GPT-4 TurboOpenAI | 1256 | 16,354 | ±7.4 | Not benchmarked | $30.00 | |
| 153 | Nova Pro 1.0Amazon | 1255 | 4,281 | ±9.2 | Not benchmarked | $3.20 | |
| 154 | Claude 3 OpusRetiredAnthropic | 1254 | 33,790 | ±6.1 | Not benchmarked | $75.00 | |
| 155 | Command R (08-2024)Cohere | 1252 | 1,639 | ±14.4 | Not benchmarked | $0.60 | |
| 156 | Claude 3.5 HaikuRetiredAnthropic | 1247 | 11,850 | ±6.2 | Not benchmarked | $4.00 | |
| 157 | Mistral Small 3Mistral AI | 1235 | 2,579 | ±11.7 | Not benchmarked | $0.08 | |
| 158 | Nova Lite 1.0Amazon | 1232 | 3,333 | ±10.4 | Not benchmarked | $0.24 | |
| 159 | Gemma 2 27BGoogle | 1228 | 12,863 | ±6.6 | Not benchmarked | $0.65 | |
| 160 | Qwen2.5 Coder 32B InstructQwen | 1221 | 956 | ±18.1 | Not benchmarked | $1.00 | |
| 161 | Nova Micro 1.0Amazon | 1210 | 3,314 | ±10.5 | Not benchmarked | $0.14 | |
| 162 | Claude 3 SonnetRetiredAnthropic | 1209 | 19,089 | ±7.6 | Not benchmarked | $15.00 | |
| 163 | Phi 4Microsoft | 1206 | 4,245 | ±9.6 | Not benchmarked | $0.14 | |
| 164 | Llama 3.1 8B InstructMeta Llama | 1193 | 8,403 | ±7.9 | Not benchmarked | $0.08 | |
| 165 | Claude 3 HaikuRetiredAnthropic | 1178 | 20,565 | ±7.2 | Not benchmarked | $1.25 | |
| 166 | GPT-4OpenAI | 1177 | 9,610 | ±9.3 | Not benchmarked | $60.00 | |
| 167 | Llama 3.2 3B InstructMeta Llama | 1106 | 1,355 | ±17.3 | Not benchmarked | $0.3325 | |
| 168 | Llama 3.2 1B InstructMeta Llama | 1055 | 1,407 | ±17.8 | Not benchmarked | $0.201 |
Only models in our catalog are listed, and each appears once with its best published run. 184 rows were read from the source on October 22, 2026, and rows that match no catalog model are left out. Every other column is a figure the board publishes for the same run.