LMArena Text Arena (Chatbot Arena) what it measures, how it is scored and who leads it

Name: LMArena Text Arena (Chatbot Arena)
Creator: LMArena (from LMSYS, UC Berkeley)
Published: 2023-05

Arenas · LMArena (from LMSYS, UC Berkeley) · introduced May 2023 · Arena rating

Scores from LMArena Text Arena, published Sep 25, 2026

Rank	Model	Arena rating
1	Claude Opus 5.5Anthropic · as “claude-opus-5.5-high”	1,509
2	Claude Opus 4.6Anthropic · as “claude-opus-4-6-high”	1,505
3	Claude Fable 5Anthropic · as “claude-fable-5-high”	1,504
4	Claude Opus 4.7Anthropic · as “claude-opus-4-7-high”	1,502
5	Claude Fable 5.1Anthropic · as “claude-fable-5.1-max”	1,501
6	Muse Spark 1.2Meta · as “muse-spark-1.2 (xHigh)”	1,496
7	Muse Spark 1.3Meta · as “muse-spark-1.3-max”	1,494
8	Gemini 3.8 FlashGoogle · as “gemini-3.8-flash-high”	1,492
9	Claude Opus 5Anthropic · as “claude-opus-5-high”	1,491
10	Muse Spark 1.1Meta · as “muse-spark-1.1”	1,491
11	Muse SparkMeta · as “muse-spark”	1,489
12	Gemini 3.7 FlashGoogle · as “gemini-3.7-flash-high”	1,488
13	Kimi K3Moonshot AI · as “kimi-k3-max”	1,488
14	Gemini 3.1 Pro PreviewGoogle · as “gemini-3.1-pro-preview”	1,487
15	Gemini 3 ProGoogle · as “gemini-3-pro”	1,485
16	GPT-5.6 SolOpenAI · as “gpt-5.6-sol-xhigh”	1,483
17	Gemini 3.6 FlashGoogle · as “gemini-3.6-flash-high”	1,482
18	GPT-5.5OpenAI · as “gpt-5.5-high”	1,481
19	Claude Opus 4.8Anthropic · as “claude-opus-4-8-high”	1,480
20	MiMo-V2.6-ProXiaomi · as “mimo-v2.6-pro”	1,480
21	GLM-5.3Z.ai · as “glm-5.3-max”	1,480
22	Qwen3.8 Max (0902)Alibaba · as “qwen3.8-max”	1,479
23	GPT-6 AstraOpenAI · as “gpt-6-astra-max”	1,478
24	Gemini 3.5 FlashGoogle · as “gemini-3.5-flash-high”	1,477
25	DeepSeek V4.1 FlashDeepSeek · as “deepseek-v4.1-flash-max”	1,477

Task: Chat: people ask anything and vote for the better of two anonymous answers.
Dataset: Millions of crowd-sourced votes on real prompts.
Method: Pairwise votes fitted with a Bradley-Terry model; style control adjusts for length and formatting.
Metric: Arena rating
Organization: LMArena (from LMSYS, UC Berkeley)
Introduced: May 2023
Paper: Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference (arXiv 2403.04132)
Official leaderboard: arena.ai/leaderboard/text

Newest first. New versions are added, never rewritten.

Version	Date	What changed
PaperMethod and statistics published.	Mar 7, 20242 years ago	Method and statistics published.
Chatbot ArenaLaunched by LMSYS.	May 20233 years ago	Launched by LMSYS.

Sources: each benchmark's paper (arXiv) and its official site or leaderboard. Scores are the boards captured here every day, exactly as published.