Skip to content

LMArena Search Arena what it measures, how it is scored and who leads it

Arenas · LMArena · Arena rating

RankModelArena rating
1GPT-5.6 SolOpenAI · as “gpt-5.6-sol-xhigh”1,257
2Claude Opus 4.6Anthropic · as “claude-opus-4-6-search”1,253
3GPT-5.5OpenAI · as “gpt-5.5-search”1,242
4Claude Opus 4.7Anthropic · as “claude-opus-4-7”1,233
5Claude Fable 5Anthropic · as “claude-fable-5-high”1,230
6Ernie 5.1Baidu · as “ernie-5.1”1,227
7Claude Sonnet 4.6Anthropic · as “claude-sonnet-4-6-search”1,221
8Grok 4.5SpaceXAI · as “grok-4.5”1,213
9Gemini 3.1 Pro GroundingGoogle · as “gemini-3.1-pro-grounding”1,210
10Gemini 3 Pro GroundingGoogle · as “gemini-3-pro-grounding”1,207
11GPT-5.2OpenAI · as “gpt-5.2-search”1,207
12Claude Opus 4.8Anthropic · as “claude-opus-4-8”1,204
13Grok 4.20 Multi AgentSpaceXAI · as “grok-4.20-multi-agent-beta-0309”1,204
14GPT-5.1OpenAI · as “gpt-5.1-search”1,199
15Gemini 3 Flash GroundingGoogle · as “gemini-3-flash-grounding”1,198
16GPT-5.4OpenAI · as “gpt-5.4-search”1,197
17Claude Sonnet 5Anthropic · as “claude-sonnet-5-search”1,194
18Grok 4.20 Beta1SpaceXAI · as “grok-4.20-beta1”1,189
19Claude Opus 4.5Anthropic · as “claude-opus-4-5-search”1,180
20GPT-5.2 NonOpenAI · as “gpt-5.2-search-non-reasoning”1,172
21Grok 4 1 FastSpaceXAI · as “grok-4-1-fast-search”1,171
22Grok 4 FastSpaceXAI · as “grok-4-fast-search”1,171
23Grok 4.3 BetaSpaceXAI · as “grok-4.3”1,165
24Claude Sonnet 4.5Anthropic · as “claude-sonnet-4-5-search”1,158
25Claude Opus 4.1Anthropic · as “claude-opus-4-1-search”1,148

About this benchmark

Task
Answer with web search grounding; people vote for the better answer.
Dataset
Crowd-sourced prompts and votes.
Method
Pairwise votes fitted with a Bradley-Terry model.
Metric
Arena rating
Organization
LMArena
Official leaderboard
arena.ai/leaderboard/search

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

New #1s by email

Saturdays, only in weeks when a leaderboard has a new #1.

Double opt-in. Unsubscribe any time.