LMArena (Chatbot Arena) Elo Leaderboard
human-preference192 models · Updated June 2026
On the LMArena (Chatbot Arena) Elo benchmark, Anthropic: Claude Opus 4.6 (Fast) ranks #1 with a score of 1500, while Llama 3.1 8B (decentralized) offers the best score-per-dollar at $0.03/1M output tokens. The full ranking, with cost per million tokens, is below.
💰 Best value
Llama 3.1 8B (decentralized) — score 1187 at $0.03/1M output tokens
| # | Model | Score | Output / 1M | Context |
|---|---|---|---|---|
| 1 | Anthropic: Claude Opus 4.6 (Fast)Anthropic | 1500 | $150.00 | 1M |
| 2 | Claude Fable 5Anthropic | 1494 | — | — |
| 3 | Claude Opus 4.7Anthropic | 1489 | — | — |
| 4 | Gemini 3.5 FlashNanoGPT | 1480 | $9.00 | 1M |
| 5 | Gemini 3.1 ProGoogle DeepMind | 1480 | — | — |
| 6 | Gemini 3 ProGoogle DeepMind | 1479 | — | — |
| 7 | Qwen3.7 Max ThinkingNanoGPT | 1475 | $7.50 | 1M |
| 8 | GPT-5.4OpenAI | 1470 | — | — |
| 9 | GLM-5.1Z.ai (Zhipu AI) | 1468 | — | — |
| 10 | GPT-5.5OpenAI | 1468 | — | — |
| 11 | ERNIE 5.1NanoGPT | 1467 | $3.00 | 119K |
| 12 | Gemini 3 Flash ThinkingNanoGPT | 1466 | $3.00 | 1M |
| 13 | Z.ai: GLM 5.2Z.ai | 1465 | $3.00 | 1M |
| 14 | Qwen3.7 Plus ThinkingNanoGPT | 1463 | $1.60 | 984K |
| 15 | Claude Opus 4.8 ThinkingNanoGPT | 1463 | $25.01 | 1M |
| 16 | MiMo-V2.5-ProXiaomi Corp | 1462 | — | — |
| 17 | Gemini 2.5 Pro (Jun 2025)Google DeepMind | 1457 | — | — |
| 18 | Claude Sonnet 4.6 ThinkingNanoGPT | 1457 | $14.99 | 1M |
| 19 | Grok 4.20 (Reasoning)xAI | 1455 | $2.50 | 1M |
| 20 | Kimi K2.6Moonshot | 1455 | — | — |
| 21 | Grok 4.20 Multi-AgentxAI | 1450 | $2.50 | 1M |
| 22 | Claude Opus 4.5Anthropic | 1450 | — | — |
| 23 | DeepSeek-V4-ProDeepSeek | 1449 | — | — |
| 24 | Qwen3.6 Max PreviewNanoGPT | 1446 | $7.80 | 246K |
| 25 | GLM-5Z.ai (Zhipu AI) | 1446 | — | — |
| 26 | Kimi K2.5Moonshot | 1445 | — | — |
| 27 | Gemma 4 31BNanoGPT | 1441 | $0.35 | 262K |
| 28 | GPT-5.1Poe | 1441 | $9.00 | 400K |
| 29 | GLM-4.6Z.ai (Zhipu AI),Tsinghua University | 1440 | — | — |
| 30 | MiniMax M3 ThinkingNanoGPT | 1440 | $1.20 | 512K |
| 31 | Qwen3.5 397B-A17BAlibaba | 1439 | — | — |
| 32 | Qwen3-Max-ThinkingAlibaba | 1439 | — | — |
| 33 | GPT-5.2OpenAI | 1438 | — | — |
| 34 | Claude Sonnet 4.5Anthropic | 1438 | — | — |
| 35 | Grok 4.1xAI | 1437 | — | — |
| 36 | Qwen3.6 PlusAlibaba Token Plan | 1437 | $0.00 | 1M |
| 37 | GLM-4.7Z.ai (Zhipu AI) | 1436 | — | — |
| 38 | MiMo-V2-ProXiaomi Corp | 1436 | — | — |
| 39 | Gemma 4 26B A4B ThinkingNanoGPT | 1435 | $0.40 | 262K |
| 40 | DeepSeek-V4-FlashDeepSeek | 1431 | — | — |
| 41 | Mistral Large 3Vercel AI Gateway | 1430 | $1.50 | 256K |
| 42 | GLM-4.5Z.ai (Zhipu AI),Tsinghua University | 1429 | — | — |
| 43 | chatgpt-4o-latest302.AI | 1429 | $15.00 | 128K |
| 44 | DeepSeek: R1 0528DeepSeek | 1428 | $2.15 | 164K |
| 45 | MiMo V2.5NanoGPT | 1427 | $0.28 | 1M |
| 46 | DeepSeek: DeepSeek V3.2DeepSeek | 1424 | $0.34 | 131K |
| 47 | MiMo V2 OmniNanoGPT | 1423 | $2.00 | 262K |
| 48 | LongCat-FlashMeituan Inc | 1422 | — | — |
| 49 | Qwen: Qwen3 VL 235B A22B ThinkingQwen | 1421 | $2.60 | 131K |
| 50 | Mistral Medium 3.5NanoGPT | 1420 | $7.50 | 256K |
| 51 | DeepSeek: DeepSeek V3.1 TerminusDeepSeek | 1420 | $0.95 | 164K |
| 52 | Qwen: Qwen3 235B A22B Thinking 2507Qwen | 1419 | $0.10 | 262K |
| 53 | DeepSeek: DeepSeek V3.1DeepSeek | 1419 | $0.79 | 164K |
| 54 | GPT-5.5 InstantZenMux | 1419 | $30.00 | 400K |
| 55 | Qwen: Qwen3 Next 80B A3B ThinkingQwen | 1419 | $0.78 | 262K |
| 56 | Claude Opus 4.1Anthropic | 1418 | — | — |
| 57 | Qwen3.5-122B-A10BAlibaba | 1418 | — | — |
| 58 | GPT-4.5OpenAI | 1417 | — | — |
| 59 | Gemini 2.5 Flash PreviewNanoGPT | 1417 | $0.60 | 1M |
| 60 | Gemini 3.1 Flash LiteNanoGPT | 1415 | $1.50 | 1M |
| 61 | Kimi K2 Thinking TurboMoonshot AI | 1414 | $8.00 | 262K |
| 62 | GPT 5.4 MiniNanoGPT | 1413 | $4.50 | 400K |
| 63 | MiMo V2 FlashNanoGPT | 1411 | $0.31 | 256K |
| 64 | grok-4-0709Jiekou.AI | 1410 | $13.50 | 256K |
| 65 | o3OpenAI | 1409 | — | — |
| 66 | Grok 4 FastRequesty | 1409 | $0.50 | 2M |
| 67 | Qwen3.5 27BNanoGPT | 1409 | $2.16 | 260K |
| 68 | Grok 4.1 FastxAI | 1408 | — | — |
| 69 | GPT-5OpenAI | 1405 | — | — |
| 70 | MiniMax M2.7NanoGPT | 1404 | $1.20 | 205K |
| 71 | Step 3.5 FlashNanoGPT | 1404 | $0.50 | 256K |
| 72 | Grok 4.3NanoGPT | 1401 | $2.50 | 1M |
| 73 | Hunyuan-T1Tencent Coding Plan (China) | 1400 | $0.00 | 131K |
| 74 | Qwen: Qwen3.5-FlashQwen | 1398 | $0.26 | 1M |
| 75 | Qwen3.5 35B A3BNanoGPT | 1396 | $1.80 | 260K |
| 76 | Claude Haiku 4.5Anthropic | 1392 | — | — |
| 77 | MiniMax-M2.1MiniMax | 1391 | — | — |
| 78 | OpenAI: GPT-5.3 ChatOpenAI | 1388 | $14.00 | 128K |
| 79 | Qwen: Qwen3 30B A3B Thinking 2507Qwen | 1384 | $0.40 | 131K |
| 80 | Z.ai: GLM 4.5 AirZ.ai | 1383 | $0.85 | 131K |
| 81 | GPT 4.1NanoGPT | 1382 | $8.00 | 1M |
| 82 | Kimi K2 0905NanoGPT | 1379 | $2.00 | 256K |
| 83 | Nemotron 3 Super 120B A12BSynthetic | 1378 | $1.00 | 262K |
| 84 | Hunyuan-TurboSTencent | 1376 | — | — |
| 85 | Claude Opus 4Anthropic | 1375 | — | — |
| 86 | DeepSeek: DeepSeek V3 0324DeepSeek | 1375 | $0.77 | 164K |
| 87 | Z.ai: GLM 4.6VZ.ai | 1375 | $0.90 | 131K |
| 88 | GPT 5.4 NanoNanoGPT | 1374 | $1.25 | 400K |
| 89 | GPT-5 miniOpenAI | 1374 | — | — |
| 90 | DeepSeek-R1 (May 2025)DeepSeek | 1373 | — | — |
| 91 | Kimi K2 0711NanoGPT | 1371 | $2.00 | 128K |
| 92 | Mistral Medium 2505routing.run | 1369 | $2.00 | 128K |
| 93 | gemini-2.5-flash-lite-preview-06-17Jiekou.AI | 1368 | $0.36 | 1M |
| 94 | Grok 3 MiniGitHub Models | 1367 | $0.00 | 128K |
| 95 | Qwen2.5-Max-2025-01-25Qiniu | 1366 | — | 128K |
| 96 | o1OpenAI | 1366 | — | — |
| 97 | Qwen3-235B-A22B-Thinking (Jul 2025)Alibaba | 1366 | — | — |
| 98 | gpt-oss-120bOpenAI | 1365 | — | — |
| 99 | Amazon: Nova 2 LiteAmazon | 1363 | $2.50 | 1M |
| 100 | MiniMax M2.5NanoGPT | 1359 | $1.20 | 205K |
| 101 | Gemma 3 27B ITNanoGPT | 1358 | $0.30 | 128K |
| 102 | Mercury 2NanoGPT | 1357 | $0.75 | 128K |
| 103 | Qwen3-Coder-480B-A35BAlibaba | 1356 | — | — |
| 104 | INTELLECT 3Cortecs | 1356 | $1.20 | 128K |
| 105 | Gemini 2.0 FlashQiniu | 1354 | — | 1M |
| 106 | Z.ai: GLM 4.7 FlashZ.ai | 1353 | $0.40 | 203K |
| 107 | OpenAI o4-mini highNanoGPT | 1353 | $4.40 | 200K |
| 108 | Step-3ZenMux | 1350 | $0.57 | 66K |
| 109 | Claude Sonnet 4Anthropic | 1348 | — | — |
| 110 | MiniMax M1NanoGPT | 1342 | $1.33 | 1M |
| 111 | MiniMax-M2MiniMax | 1342 | — | — |
| 112 | Trinity Large ThinkingNanoGPT | 1342 | $0.90 | 262K |
| 113 | GPT 4.1 MiniNanoGPT | 1340 | $1.60 | 1M |
| 114 | Qwen: Qwen3 32BQwen | 1340 | $0.28 | 131K |
| 115 | NVIDIA: Llama 3.3 Nemotron Super 49B V1.5NVIDIA | 1338 | $0.40 | 131K |
| 116 | o3-miniOpenAI | 1337 | — | — |
| 117 | Gemma 3 12B ITNanoGPT | 1334 | $0.27 | 128K |
| 118 | GLM 4.5V ThinkingNanoGPT | 1333 | $1.80 | 64K |
| 119 | DeepSeek-V3 (Mar 2025)DeepSeek | 1333 | — | — |
| 120 | Cohere Command A (08/2025)NanoGPT | 1331 | $10.00 | 256K |
| 121 | GLM 4 Plus 0111NanoGPT | 1331 | $10.00 | 128K |
| 122 | QwQ-32BAlibaba | 1329 | — | — |
| 123 | GPT-5 nanoOpenAI | 1320 | — | — |
| 124 | Llama-3.1-Nemotron-Ultra-253B-v1Nebius Token Factory | 1319 | $1.80 | 128K |
| 125 | Gemini 1.5 ProGoogle DeepMind | 1319 | — | — |
| 126 | o1-miniOpenAI | 1317 | — | — |
| 127 | Qwen: Qwen3 30B A3BQwen | 1317 | $0.50 | 131K |
| 128 | Claude 3.7 SonnetAnthropic | 1314 | — | — |
| 129 | Google: Gemma 3n 4BGoogle | 1306 | $0.12 | 33K |
| 130 | Grok-2xAI | 1305 | — | — |
| 131 | Yi-Lightning01.AI | 1302 | — | — |
| 132 | GPT-4o (Mar 2025)OpenAI | 1301 | — | — |
| 133 | Olmo 3 32B ThinkNanoGPT | 1299 | $0.45 | 128K |
| 134 | Claude 3.5 SonnetAnthropic | 1298 | — | — |
| 135 | Granite 4.1 8BNanoGPT | 1293 | $0.10 | 131K |
| 136 | Gemma 3 4B ITNanoGPT | 1291 | $0.20 | 128K |
| 137 | GLM-4-PlusZ.ai (Zhipu AI) | 1290 | — | — |
| 138 | Hunyuan-LargeTencent | 1288 | — | — |
| 139 | Llama 4 Maverick 17B 128E InstructDigitalOcean | 1288 | $0.87 | 1M |
| 140 | gpt-oss-20bOpenAI | 1288 | — | — |
| 141 | Gemini 1.5 FlashNanoGPT | 1287 | $0.31 | 2M |
| 142 | GPT-4o miniOpenAI | 1287 | — | — |
| 143 | GPT 4.1 NanoNanoGPT | 1285 | $0.40 | 1M |
| 144 | Llama-3.1-Nemotron-70B-InstructNVIDIA,Meta AI | 1283 | — | — |
| 145 | MercuryInception Labs | 1282 | — | — |
| 146 | Llama 4 Scout 17B 16E InstructCloudflare Workers AI | 1281 | $0.85 | 131K |
| 147 | Llama 3.3 70BMeta AI | 1275 | — | — |
| 148 | GPT-4 Turbo (Apr 2024)OpenAI | 1272 | — | — |
| 149 | DeepSeek-V2.5DeepSeek | 1271 | — | — |
| 150 | Qwen2.5-72BAlibaba | 1269 | — | — |
| 151 | Mistral Large 2407 | 1266 | $6.00 | 131K |
| 152 | Mistral Large 2411NanoGPT | 1265 | $6.00 | 128K |
| 153 | Claude 3 OpusAnthropic | 1262 | — | — |
| 154 | Meta: Llama 3.1 70B InstructMeta | 1261 | $0.40 | 131K |
| 155 | Claude 3.5 HaikuQiniu | 1255 | — | 200K |
| 156 | Reka CoreReka AI | 1248 | — | — |
| 157 | Jamba 1.5-LargeAI21 Labs | 1237 | — | — |
| 158 | Google: Gemma 2 27BGoogle | 1231 | $0.65 | 8K |
| 159 | Qwen2.5 Coder 32B Instruct | 1230 | $1.00 | 128K |
| 160 | Cohere: Command R+ (08-2024)Cohere | 1229 | $10.00 | 128K |
| 161 | GLM-4 (0520)Z.ai (Zhipu AI) | 1226 | — | — |
| 162 | Nemotron-4 340BNVIDIA | 1225 | — | — |
| 163 | Aya Expanse 32BCohere | 1224 | — | 128K |
| 164 | Llama 3-70BMeta AI | 1221 | — | — |
| 165 | Claude 3 SonnetAnthropic | 1218 | — | — |
| 166 | Qwen FlashAlibaba (China) | 1218 | $0.22 | 1M |
| 167 | Phi-4Azure | 1217 | $0.50 | 128K |
| 168 | Qwen2-72BAlibaba | 1203 | — | — |
| 169 | Anthropic: Claude 3 HaikuAnthropic | 1195 | $1.25 | 200K |
| 170 | Cohere: Command RNanoGPT | 1187 | $1.43 | 128K |
| 171 | AI21 Jamba 1.5 MiniGitHub Models | 1187 | $0.00 | 256K |
| 172 | Llama 3.1 8B (decentralized)NanoGPT | 1187 | $0.03 | 128K |
| 173 | Aya Expanse 8BCohere | 1185 | — | 8K |
| 174 | Qwen1.5-72BAlibaba | 1166 | — | — |
| 175 | Meta: Llama 3 8B InstructMeta | 1166 | $0.14 | 8K |
| 176 | Gemma 2 2b ItNvidia | 1156 | $0.00 | 128K |
| 177 | Mixtral 8x7B Instruct v0.1Cortecs | 1132 | $0.68 | 32K |
| 178 | Google Gemini Pro Latest | 1131 | $12.00 | 1M |
| 179 | Yi-34B01.AI | 1129 | — | — |
| 180 | GPT-3.5 Turbo 0125Azure | 1125 | $1.50 | 16K |
| 181 | DBRXDatabricks | 1119 | — | — |
| 182 | Llama 2-70BMeta AI | 1115 | — | — |
| 183 | Phi-3-small instruct (128k)GitHub Models | 1110 | $0.00 | 128K |
| 184 | Llama 3.2 3b InstructNanoGPT | 1110 | $0.05 | 131K |
| 185 | GPT-3.5 Turbo 1106Azure | 1094 | $2.00 | 16K |
| 186 | Meta: Llama 3.2 1B InstructMeta | 1055 | $0.20 | 131K |
| 187 | Falcon-180BTechnology Innovation Institute | 1054 | — | — |
| 188 | Llama 2-7BMeta AI | 1053 | — | — |
| 189 | Phi-3-mini instruct (128k)GitHub Models | 1050 | $0.00 | 128K |
| 190 | PaLM 2Google | 1027 | — | — |
| 191 | Mistral 7BMistral | 1024 | $0.25 | 8K |
| 192 | ChatGLM3-6BZ.ai (Zhipu AI) | 972 | — | — |
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
Pricing is indicative — confirm with the provider before production use. Updated June 2026.