Arena Multi-turn Elo

Name: Arena Multi-turn Elo
Creator: KnowAI Space
License: https://www.gnu.org/licenses/gpl-3.0.html

About This Benchmark

Arena Elo score from user preference votes on multi-turn conversation prompts.

Rank	Model
#1	Anthropic Claude Opus 4.6	1518
#2	Grok Grok 4.20	1513
#3	Grok Grok 4.20 (Reasoning)	1513
#4	Google Gemini 3.1 Pro	1504
#5	ChatGPT GPT-5.4	1500
#6	ChatGPT GPT-5.4 Pro	1500
#7	Anthropic Claude Opus 4.5	1487
#8	Google Gemini 3 Flash	1483
#9	Z.ai GLM-5.1	1479
#10	Anthropic Claude Sonnet 4.6	1478
#11	ChatGPT GPT-5.4 Mini	1478
#12	Anthropic Claude Opus 4.1	1473
#13	Anthropic Claude Sonnet 4.5	1468
#14	Google Gemma 4 31B	1463
#15	Z.ai GLM-5	1461
#16	Xiaomi MiMo-V2-Pro	1459
#17	Moonshot AI Kimi K2.5	1455
#18	Google Gemini 2.5 Pro	1452
#19	Qwen Qwen3.5 397B A17B	1449
#20	Google Gemini 3.1 Flash Lite	1447
#21	Baidu ERNIE 5.0 Thinking	1441
#22	Anthropic Claude Opus 4	1437
#23	Meituan Longcat Flash Chat	1426
#24	DeepSeek DeepSeek V3.2	1426
#25	Anthropic Claude Haiku 4.5	1421
#26	ChatGPT GPT-5	1420
#27	Anthropic Claude Sonnet 4	1419
#28	Grok Grok 4.1 Fast	1416
#29	Grok Grok 4.1 Fast (Reasoning)	1416
#30	MiniMax MiniMax M2.5	1413
#31	ChatGPT GPT-5.4 Nano	1412
#32	MiniMax MiniMax M2.7	1407
#33	Google Gemini 2.5 Flash	1403
#34	Google Gemini 2.5 Flash Lite	1374
#35	Arcee AI Trinity Large Thinking	1372
#36	ChatGPT GPT-5 Mini	1372
#37	ChatGPT GPT OSS 120B	1328
#38	Meta AI Llama 4 Maverick	1324
#39	Amazon Nova 2 Lite	1324
#40	ChatGPT GPT-5 Nano	1323
#41	Meta AI Llama 4 Scout	1320
#42	ChatGPT GPT-4.1	1298