LMArena (Chatbot Arena) Elo Leaderboard
human-preference245 models · Updated October 2026
On the LMArena (Chatbot Arena) Elo benchmark, Gemini 4 Carbon ranks #1 with a score of 1540, while Gemma 3 4B IT offers the best score-per-dollar at $0.05/1M output tokens. The full ranking, with cost per million tokens, is below.
💰 Best value
Gemma 3 4B IT — score 1291 at $0.05/1M output tokens
| # | Model | Score | Output / 1M | Context |
|---|---|---|---|---|
| 1 | Gemini 4 CarbonGoogle DeepMind | 1540 | — | — |
| 2 | Gemini 4 ArgonGoogle | 1529 | $10.00 | 1M |
| 3 | Claude Opus 5.5Anthropic | 1516 | $20.00 | 1M |
| 4 | Claude Fable 5.1Anthropic | 1514 | $50.00 | 1M |
| 5 | Claude Opus 5Anthropic | 1508 | $25.00 | 1M |
| 6 | Claude Opus 4.6Anthropic | 1502 | $25.00 | 1M |
| 7 | Muse WonderMeta | 1499 | $20.00 | 1M |
| 8 | Gemini 3.8 FlashGoogle | 1498 | $4.13 | 1M |
| 9 | Claude Fable 5Anthropic | 1492 | $50.00 | 1M |
| 10 | MiMo-V2.6-ProXiaomi | 1491 | $0.87 | 1.1M |
| 11 | Muse Spark 1.3Meta | 1491 | $4.25 | 1M |
| 12 | Claude Opus 4.7Anthropic | 1490 | $25.00 | 1M |
| 13 | Gemini 3.7 FlashGoogle | 1489 | $3.75 | 1M |
| 14 | Qwen3.8 MaxAlibaba | 1483 | $6.00 | 1M |
| 15 | Muse Spark 1.2Meta | 1483 | $4.25 | 1M |
| 16 | Gemini 3.5 FlashGoogle | 1482 | $9.00 | 1M |
| 17 | Gemini 3.1 Pro PreviewGoogle | 1480 | $12.00 | 1M |
| 18 | Gemini 3.6 FlashGoogle | 1480 | $3.75 | 1M |
| 19 | Gemini 3 ProOpenCode | 1479 | $12.00 | 1M |
| 20 | Muse Spark 1.1Meta | 1478 | $4.25 | 1M |
| 21 | Kimi K3MoonshotAI | 1476 | $15.00 | 1M |
| 22 | Muse SparkMeta AI | 1474 | — | — |
| 23 | GLM 5.3Z.ai | 1474 | $3.60 | 1M |
| 24 | GLM 5.3 FlashZ.ai | 1473 | $0.07 | 1M |
| 25 | Claude Sonnet 5.5Anthropic | 1473 | $10.00 | 1M |
| 26 | GPT-5.5OpenAI | 1472 | $30.00 | 1.1M |
| 27 | GLM 5.2Z.ai | 1470 | $7.00 | 1M |
| 28 | GPT-5.4OpenAI | 1469 | $15.00 | 1.1M |
| 29 | ERNIE 5.1NanoGPT | 1468 | $3.00 | 119K |
| 30 | Gemini 3 FlashOpenCode | 1466 | $3.00 | 1M |
| 31 | DeepSeek V4.1 FlashDeepSeek | 1465 | $0.50 | 1M |
| 32 | MiMo-V2.5-ProXiaomi | 1464 | $0.87 | 1.1M |
| 33 | GLM 5.1Z.ai | 1463 | $4.40 | 205K |
| 34 | Claude Opus 4.8Anthropic | 1461 | $25.00 | 1M |
| 35 | Claude Sonnet 4.6Anthropic | 1458 | $15.00 | 1M |
| 36 | Gemini 2.5 ProGoogle | 1458 | $10.00 | 1M |
| 37 | GPT-5.6 SolOpenAI | 1457 | $20.00 | 1.1M |
| 38 | Kimi K2.6MoonshotAI | 1456 | $3.41 | 262K |
| 39 | Qwen3.7 PlusAlibaba | 1455 | $1.60 | 1M |
| 40 | MiMo-V2.6-FlashXiaomi | 1451 | $0.28 | 1.1M |
| 41 | DeepSeek V4 Pro 0423DeepSeek | 1451 | $1.63 | 1M |
| 42 | grok-4.20-beta-0309-reasoning302.AI | 1451 | $6.00 | 2M |
| 43 | Claude Opus 4.5Anthropic | 1451 | $25.00 | 200K |
| 44 | grok-4.20-multi-agent-beta-0309302.AI | 1451 | $6.00 | 2M |
| 45 | GPT-5.6 TerraOpenAI | 1448 | $12.00 | 1.1M |
| 46 | Grok 4.5SpaceXAI | 1448 | $6.00 | 500K |
| 47 | Qwen3.6 Max PreviewAlibaba | 1447 | $7.80 | 262K |
| 48 | GLM 5Z.ai | 1446 | $3.20 | 205K |
| 49 | GPT-6.1 SolOpenAI | 1446 | $10.00 | 1.1M |
| 50 | Nemotron 3 UltraNVIDIA | 1445 | $2.40 | 262K |
| 51 | Kimi K2.5MoonshotAI | 1445 | $2.50 | 262K |
| 52 | ERNIE 5.0ZenMux | 1445 | $3.37 | 128K |
| 53 | Gemma 4 31BGoogle | 1443 | $0.40 | 262K |
| 54 | Step 5 PreviewStepFun | 1443 | $2.70 | 1M |
| 55 | GPT-5.1OpenAI | 1443 | $10.00 | 400K |
| 56 | Claude Sonnet 5Anthropic | 1443 | $10.00 | 1M |
| 57 | InklingThinking Machines | 1443 | $4.05 | 524K |
| 58 | GPT-6 AstraOpenAI | 1441 | $50.00 | 1.1M |
| 59 | GLM 4.6Z.ai | 1440 | $2.00 | 205K |
| 60 | Hy3Tencent | 1439 | $0.53 | 262K |
| 61 | Qwen3-Max-PreviewiFlow | 1439 | $0.00 | 256K |
| 62 | Claude Sonnet 4.5Anthropic | 1439 | $15.00 | 1M |
| 63 | GPT-5.2 ChatOpenAI | 1439 | $14.00 | 128K |
| 64 | Qwen3.8 27BQwen | 1438 | $1.88 | 1M |
| 65 | GLM 5V TurboZ.ai | 1438 | $4.00 | 203K |
| 66 | Qwen3.5 397B-A17BAlibaba | 1438 | $3.60 | 262K |
| 67 | grok-4.1302.AI | 1436 | $10.00 | 200K |
| 68 | Qwen3.6 PlusQwen | 1436 | $1.95 | 1M |
| 69 | MiMo-V2-ProXiaomi | 1436 | $0.87 | 1M |
| 70 | Gemini 3.5 Flash LiteGoogle | 1435 | $2.50 | 1M |
| 71 | GLM 4.7Z.ai | 1435 | $1.98 | 205K |
| 72 | Gemma 4 26B A4BGoogle | 1434 | $0.30 | 262K |
| 73 | DeepSeek V4 Flash 0423DeepSeek | 1432 | $0.18 | 1M |
| 74 | MiniMax M3MiniMax | 1431 | $1.20 | 1M |
| 75 | GPT-5.6 LunaOpenAI | 1431 | $1.20 | 1.1M |
| 76 | Mistral Large 4Mistral | 1430 | $2.10 | 1M |
| 77 | GLM 4.5Z.ai | 1430 | $2.20 | 131K |
| 78 | Grok 4.6SpaceXAI | 1429 | $6.00 | 500K |
| 79 | chatgpt-4o-latest302.AI | 1428 | $15.00 | 128K |
| 80 | Mistral Large 3 2512Mistral | 1428 | $1.50 | 262K |
| 81 | MiMo-V2.5Xiaomi | 1428 | $0.28 | 1.1M |
| 82 | DeepSeek-R1Azure | 1428 | $5.40 | 164K |
| 83 | Beam 1.0 501B A23BReflection | 1425 | $1.00 | 1M |
| 84 | DeepSeek V3.2 ExpDeepSeek | 1425 | $0.41 | 164K |
| 85 | DeepSeek V3.2DeepSeek | 1424 | $0.42 | 164K |
| 86 | Mistral Medium 3.1Mistral | 1424 | $2.00 | 131K |
| 87 | MiMo-V2-OmniXiaomi | 1423 | $0.28 | 262K |
| 88 | LongCat Flash ChatVercel AI Gateway | 1423 | — | 128K |
| 89 | Mistral Medium 3.5Mistral | 1421 | $7.50 | 262K |
| 90 | Qwen3 VL 235B A22B InstructQwen | 1420 | $1.90 | 262K |
| 91 | DeepSeek V3.1DeepSeek | 1420 | $0.95 | 164K |
| 92 | Claude Opus 4.1Anthropic | 1419 | $75.00 | 200K |
| 93 | GPT-5.5 InstantZenMux | 1419 | $30.00 | 400K |
| 94 | Qwen3 235B A22B Instruct 2507Qwen | 1419 | $0.55 | 262K |
| 95 | DeepSeek V3.1 TerminusDeepSeek | 1419 | $1.00 | 164K |
| 96 | Qwen3.5 122B-A10BAlibaba | 1418 | $3.20 | 262K |
| 97 | Qwen3-Next 80B-A3B InstructAlibaba | 1417 | $2.00 | 131K |
| 98 | Gemini 2.5 FlashGoogle | 1417 | $2.50 | 1M |
| 99 | GPT-5.2OpenAI | 1416 | $14.00 | 400K |
| 100 | Gemini 3.1 Flash Lite PreviewGoogle | 1416 | $1.50 | 1M |
| 101 | Qwen3 235B-A22BAlibaba | 1415 | $2.80 | 131K |
| 102 | Kimi K2 Thinking TurboMoonshot AI | 1414 | $8.00 | 262K |
| 103 | Inkling SmallThinking Machines | 1414 | $1.20 | 524K |
| 104 | Qwen3 MaxAlibaba | 1413 | $6.00 | 262K |
| 105 | GPT-5.4 MiniOpenAI | 1412 | $4.50 | 400K |
| 106 | Grok-4Poe | 1411 | $15.00 | 256K |
| 107 | MiMo-V2-FlashXiaomi | 1411 | $0.28 | 262K |
| 108 | o3OpenAI | 1410 | $8.00 | 200K |
| 109 | Qwen3.5 27BAlibaba | 1409 | $2.40 | 262K |
| 110 | Grok 4.1 FastOfox | 1408 | $0.50 | 2M |
| 111 | Grok 4 FastZenMux | 1407 | $0.50 | 2M |
| 112 | gemini-2.5-flash-preview-05-20Jiekou.AI | 1406 | $3.33 | 1M |
| 113 | GPT-5OpenAI | 1406 | $10.00 | 400K |
| 114 | MiniMax M2.7MiniMax | 1405 | $1.20 | 205K |
| 115 | GPT-5 ChatOpenAI | 1403 | $10.00 | 128K |
| 116 | Step 3.5 FlashStepFun | 1403 | $0.30 | 262K |
| 117 | Hy3 previewTencent TokenHub | 1401 | $0.00 | 256K |
| 118 | Claude Haiku 5.5Anthropic | 1401 | $0.50 | 1M |
| 119 | Hunyuan-T1Tencent Coding Plan (China) | 1401 | $0.00 | 131K |
| 120 | Qwen3-VL 235B-A22BAlibaba | 1400 | $2.80 | 131K |
| 121 | Grok 4.7SpaceXAI | 1398 | $6.00 | 500K |
| 122 | GPT-6 SolOpenAI | 1397 | $10.00 | 1.1M |
| 123 | Qwen3.5 FlashAlibaba | 1397 | $0.40 | 1M |
| 124 | Grok 4.3SpaceXAI | 1397 | $2.50 | 1M |
| 125 | Claude Haiku 4.5Anthropic | 1397 | $5.00 | 200K |
| 126 | Qwen3.5 35B-A3BAlibaba | 1395 | $2.00 | 262K |
| 127 | GPT-6 LunaOpenAI | 1389 | $0.50 | 1.1M |
| 128 | GPT-5.3 ChatOpenAI | 1389 | $14.00 | 128K |
| 129 | Solar Pro 4Upstage | 1389 | $1.20 | 524K |
| 130 | Qwen3 30B A3B Instruct 2507Qwen | 1384 | $0.30 | 262K |
| 131 | GLM 4.5 AirZ.ai | 1384 | $0.85 | 131K |
| 132 | GPT-4.1OpenAI | 1383 | $8.00 | 1M |
| 133 | gemini-2.5-flash-lite-preview-06-17Jiekou.AI | 1379 | $0.36 | 1M |
| 134 | Nemotron 3 SuperNVIDIA | 1377 | $2.40 | 1M |
| 135 | Hunyuan-TurboSTencent Coding Plan (China) | 1377 | $0.00 | 131K |
| 136 | Claude Opus 4Anthropic | 1377 | $75.00 | 200K |
| 137 | GLM 4.6VZ.ai | 1376 | $0.90 | 131K |
| 138 | DeepSeek V3DeepSeek | 1375 | $1.03 | 164K |
| 139 | GPT-5 MiniOpenAI | 1373 | $2.00 | 400K |
| 140 | GPT-5.4 NanoOpenAI | 1372 | $1.25 | 400K |
| 141 | Mistral Medium 3Mistral | 1370 | $2.00 | 131K |
| 142 | Grok 3 MiniPoe | 1370 | $0.50 | 131K |
| 143 | Qwen3 Next 80B A3B ThinkingQwen | 1368 | $1.20 | 262K |
| 144 | Qwen 2.5 MaxNanoGPT | 1367 | $6.39 | 32K |
| 145 | o1OpenAI | 1366 | $60.00 | 200K |
| 146 | gpt-oss-120b (batch)OpenAI | 1365 | $0.14 | 131K |
| 147 | Amazon Nova 2 LiteNanoGPT | 1362 | $4.25 | 1M |
| 148 | inclusionAI/Ling-flash-2.0SiliconFlow | 1361 | $0.57 | 131K |
| 149 | MiniMax M2.5MiniMax | 1359 | $1.20 | 205K |
| 150 | Gemma 3 27B ITCortecs | 1358 | $0.30 | 131K |
| 151 | Qwen3-Coder 480B-A35B InstructAlibaba | 1357 | $7.50 | 262K |
| 152 | INTELLECT 3Cortecs | 1356 | $1.20 | 128K |
| 153 | Mercury 2Inception | 1355 | $0.75 | 128K |
| 154 | Gemini 2.0 FlashGoogle | 1354 | $0.40 | 1M |
| 155 | o4 MiniOpenAI | 1353 | $4.40 | 200K |
| 156 | o1-previewOpenAI | 1353 | — | — |
| 157 | Claude Sonnet 4 (latest)Anthropic | 1351 | $15.00 | 200K |
| 158 | GLM 4.7 FlashZ.ai | 1351 | $0.40 | 200K |
| 159 | Step-3NanoGPT | 1350 | $0.65 | 66K |
| 160 | Nemotron 3 Nano 30B A3BNVIDIA | 1348 | $0.24 | 262K |
| 161 | MiniMax M1MiniMax | 1343 | $2.20 | 1M |
| 162 | GPT-4.1 MiniOpenAI | 1340 | $1.60 | 1M |
| 163 | Trinity Large ThinkingArcee AI | 1340 | $0.80 | 262K |
| 164 | MiniMax M2MiniMax | 1340 | $1.02 | 205K |
| 165 | Qwen3 32BQwen | 1340 | $0.28 | 131K |
| 166 | Trinity Large PreviewOpenCode | 1339 | $0.00 | 131K |
| 167 | Llama 3.3 Nemotron Super 49B V1.5NVIDIA | 1338 | $0.40 | 131K |
| 168 | Mistral Small 3.2Mistral | 1338 | $0.30 | 128K |
| 169 | o3 MiniOpenAI | 1337 | $4.40 | 200K |
| 170 | Gemma 3 12B ITKilo Gateway | 1334 | $0.12 | 131K |
| 171 | GLM 4.5VZ.ai | 1333 | $1.80 | 66K |
| 172 | Command ACohere | 1331 | $10.00 | 256K |
| 173 | GLM-4 PlusNanoGPT | 1331 | $7.50 | 128K |
| 174 | QwQ 32BAlibaba (China) | 1329 | $0.86 | 131K |
| 175 | Qwen-PlusQwen | 1326 | $0.78 | 1M |
| 176 | Step-2 16k ExpNanoGPT | 1321 | $19.99 | 16K |
| 177 | GPT-5 NanoOpenAI | 1320 | $0.40 | 400K |
| 178 | Llama-3.1-Nemotron-Ultra-253B-v1Nebius Token Factory | 1320 | $1.80 | 128K |
| 179 | Gemini 1.5 ProGoogle DeepMind | 1319 | — | — |
| 180 | Granite 4.2 8BIBM | 1319 | $0.15 | 131K |
| 181 | o1-miniAzure | 1317 | $4.40 | 128K |
| 182 | Qwen3 30B A3BQwen | 1317 | $0.50 | 41K |
| 183 | Claude 3.7 SonnetHelicone | 1314 | $15.00 | 200K |
| 184 | Hunyuan TurboTencent | 1311 | — | — |
| 185 | Llama 3.3 Nemotron Super 49B v1Nvidia | 1308 | $0.00 | 131K |
| 186 | Gemma 3n E4b ItNvidia | 1306 | $0.00 | 128K |
| 187 | Grok-2xAI | 1305 | — | — |
| 188 | Yi LightningNanoGPT | 1302 | $0.20 | 12K |
| 189 | GPT-4oOpenAI | 1300 | $10.00 | 128K |
| 190 | Olmo 3 32B ThinkAllenAI | 1299 | $0.50 | 66K |
| 191 | Claude 3.5 SonnetDigitalOcean | 1298 | $15.00 | 200K |
| 192 | DeepSeek-V2.5DeepSeek | 1294 | — | — |
| 193 | Gemma 3 4B ITKilo Gateway | 1291 | $0.05 | 131K |
| 194 | Granite 4.1 8BIBM | 1291 | $0.10 | 131K |
| 195 | Hunyuan-LargeTencent | 1288 | — | — |
| 196 | gpt-oss-20b (batch)OpenAI | 1287 | $0.11 | 131K |
| 197 | Llama 4 Maverick 17B 128E InstructVertex | 1287 | $1.15 | 524K |
| 198 | Gemini 1.5 FlashNanoGPT | 1287 | $0.31 | 2M |
| 199 | GPT-4o-miniOpenAI | 1286 | $0.60 | 128K |
| 200 | GPT-4.1 NanoOpenAI | 1285 | $0.40 | 1M |
| 201 | Llama 3.1 405B InstructCortecs | 1284 | $0.00 | 128K |
| 202 | Llama 3.1 Nemotron 70B InstructNvidia | 1283 | $0.00 | 128K |
| 203 | MercuryInception Labs | 1282 | — | — |
| 204 | Qwen MaxAlibaba | 1282 | $6.40 | 33K |
| 205 | Llama 4 Scout 17B 16E InstructAzure | 1279 | $0.78 | 128K |
| 206 | Mistral Small 31 24b InstructNanoGPT | 1277 | $0.30 | 128K |
| 207 | Llama 3.3 70B InstructMeta | 1274 | $0.32 | 131K |
| 208 | GPT-4 TurboOpenAI | 1272 | $30.00 | 128K |
| 209 | Qwen2.5 72B InstructAlibaba | 1269 | $5.60 | 131K |
| 210 | Mistral Large 2Mistral AI | 1266 | — | — |
| 211 | Mistral LargeMistral AI | 1265 | $6.00 | 128K |
| 212 | GPT-4 Turbo PreviewOpenAI | 1263 | $30.00 | 128K |
| 213 | Claude 3 OpusLLM Gateway | 1262 | $75.00 | 200K |
| 214 | Llama 3.1 70B InstructMeta | 1261 | $0.40 | 131K |
| 215 | Amazon Nova Pro 1.0NanoGPT | 1259 | $3.20 | 300K |
| 216 | Claude 3.5 HaikuHelicone | 1255 | $4.00 | 200K |
| 217 | Magistral Medium (latest)Mistral | 1255 | $5.00 | 262K |
| 218 | Reka CoreReka AI | 1248 | — | — |
| 219 | Granite-4.0-H-Smallwatsonx.ai | 1241 | $0.27 | 131K |
| 220 | Jamba 1.5-LargeAI21 Labs | 1237 | — | — |
| 221 | Qwen2.5 Coder 32B InstructAlibaba (China) | 1230 | $1.00 | 33K |
| 222 | Command R (08-2024)Cohere | 1229 | $0.60 | 128K |
| 223 | Amazon Nova Lite 1.0NanoGPT | 1228 | $0.24 | 300K |
| 224 | GLM-4 (0116)Z.ai (Zhipu AI) | 1226 | — | — |
| 225 | Aya Expanse 32BCohere | 1224 | $1.50 | 128K |
| 226 | Llama 3 70B InstructDevPass (LLM Gateway) | 1221 | $0.74 | 8K |
| 227 | Claude 3 SonnetAnthropic | 1218 | — | — |
| 228 | Phi 4Microsoft | 1217 | $0.14 | 16K |
| 229 | Amazon Nova Micro 1.0NanoGPT | 1208 | $0.14 | 128K |
| 230 | GPT-4OpenAI | 1206 | $60.00 | 8K |
| 231 | Claude 3 HaikuAnthropic | 1195 | $1.25 | 200K |
| 232 | Ministral 8B (latest)Mistral | 1191 | $0.15 | 262K |
| 233 | Llama 3.1 8B InstructMeta | 1187 | $0.08 | 131K |
| 234 | Aya Expanse 8BCohere | 1185 | — | 8K |
| 235 | Llama 3 8B InstructMeta | 1166 | $0.14 | 8K |
| 236 | Mistral Medium (latest)Mistral | 1165 | $7.50 | 262K |
| 237 | Gemma 2 2b ItNvidia | 1156 | $0.00 | 128K |
| 238 | Mixtral 8x7B Instruct v0.1Cortecs | 1132 | $0.76 | 32K |
| 239 | GPT-3.5 TurboOpenAI | 1125 | $1.50 | 16K |
| 240 | Llama 3.2 3B InstructMeta | 1110 | $0.33 | 131K |
| 241 | mistral-7b-instruct-v0.2Cortecs | 1090 | $0.22 | 32K |
| 242 | Llama 3.2 1B InstructMeta | 1055 | $0.20 | 60K |
| 243 | PaLM 2Google | 1027 | — | — |
| 244 | ChatGLM3-6BZ.ai (Zhipu AI) | 972 | — | — |
| 245 | LLaMA-13BMeta AI | 833 | — | — |
What does LMArena (Chatbot Arena) Elo test?
Overview This dataset contains ALL in-the-wild conversation crowdsourced from Search Arena between March 18, 2025 and May 8, 2025. It includes 24,069 multi-turn conversations with search-LLMs across diverse intents, languages, and topics—alongs...
Frequently asked questions
What is the LMArena (Chatbot Arena) Elo benchmark?
Overview This dataset contains ALL in-the-wild conversation crowdsourced from Search Arena between March 18, 2025 and May 8, 2025. It includes 24,069 multi-turn conversations with search-LLMs across diverse intents, languages, and topics—alongs...
Which model is best on LMArena (Chatbot Arena) Elo?
Gemini 4 Carbon (Google DeepMind) currently leads the LMArena (Chatbot Arena) Elo leaderboard with a score of 1540.
What is the best value model on LMArena (Chatbot Arena) Elo?
Gemma 3 4B IT offers the best score-per-dollar on LMArena (Chatbot Arena) Elo — a score of 1291 at $0.05 per 1M output tokens.
Pricing is indicative — confirm with the provider before production use. Updated October 2026.