Model Markets
Get Free API KeyAPI

LMArena Search Arena

LMArena · search-2026-10-13-a64ca3b12b6e · Blind human preferences for search-augmented responses in LMArena.

Model Ranking

Source
#1GPT-5.6 SolOpenAI1,261.13rating—60,437gpt-5.6-sol-xhigh; 95% CI [1254.84060485, 1267.42091382]; votes 60437; rank 1unknownOriginal sourcewinner eligibleThird-party benchmark
#2Qwen3.8 MaxQwen1,261.08rating—1,870qwen3.8-max; 95% CI [1247.89214823, 1274.26022415]; votes 1870; rank 2unknownOriginal sourcewinner eligibleThird-party benchmark
#3Claude Opus 4.6Anthropic1,251.37rating—173,191claude-opus-4-6-search; 95% CI [1246.82915739, 1255.90736955]; votes 173191; rank 3unknownOriginal sourcewinner eligibleThird-party benchmark
#4GPT-5.5OpenAI1,246.55rating—125,957gpt-5.5-search; 95% CI [1241.65410036, 1251.45490149]; votes 125957; rank 4unknownOriginal sourcewinner eligibleThird-party benchmark
#5Claude Fable 5.1Anthropic1,245.85rating—12,782claude-fable-5.1-high; 95% CI [1238.79268797, 1252.89817971]; votes 12782; rank 5unknownOriginal sourcewinner eligibleThird-party benchmark
#6Claude Opus 5Anthropic1,241.57rating—13,285claude-opus-5-high; 95% CI [1234.69644002, 1248.43868868]; votes 13285; rank 6unknownOriginal sourcewinner eligibleThird-party benchmark
#7GPT-6 AstraOpenAI1,233.62rating—18,483gpt-6-astra-max; 95% CI [1225.77363708, 1241.45798456]; votes 18483; rank 7unknownOriginal sourcewinner eligibleThird-party benchmark
#8GPT-6.1 SolOpenAI1,233.10rating—5,555gpt-6.1-sol-max; 95% CI [1224.24145466, 1241.96324521]; votes 5555; rank 8unknownOriginal sourcewinner eligibleThird-party benchmark
#9Claude Opus 4.7Anthropic1,230.15rating—127,506claude-opus-4-7; 95% CI [1225.28137992, 1235.01180806]; votes 127506; rank 9unknownOriginal sourcewinner eligibleThird-party benchmark
#10Claude Fable 5Anthropic1,229.14rating—73,596claude-fable-5-high; 95% CI [1222.63001070, 1235.64949261]; votes 73596; rank 10unknownOriginal sourcewinner eligibleThird-party benchmark
#11ERNIE 5.1Baidu1,226.72rating—4,024ernie-5.1; 95% CI [1217.24341628, 1236.18668968]; votes 4024; rank 11unknownOriginal sourcewinner eligibleThird-party benchmark
#12Claude Sonnet 4.6Anthropic1,220.50rating—173,986claude-sonnet-4-6-search; 95% CI [1216.03052081, 1224.97904735]; votes 173986; rank 12unknownOriginal sourcewinner eligibleThird-party benchmark
#13Grok 4.5xAI1,215.94rating—76,458grok-4.5; 95% CI [1210.07957437, 1221.79459071]; votes 76458; rank 13unknownOriginal sourcewinner eligibleThird-party benchmark
#14Claude Opus 5.5Anthropic1,212.75rating—10,842claude-opus-5.5-medium; 95% CI [1205.66309681, 1219.84195206]; votes 10842; rank 14unknownOriginal sourcewinner eligibleThird-party benchmark
#15Gemini 3.1 ProGoogle DeepMind1,208.81rating—151,436gemini-3.1-pro-grounding; 95% CI [1204.12797055, 1213.48336905]; votes 151436; rank 15unknownOriginal sourcewinner eligibleThird-party benchmark
#16GPT-5.2OpenAI1,207.24rating—55,634gpt-5.2-search; 95% CI [1201.69874997, 1212.78027063]; votes 55634; rank 16unknownOriginal sourcewinner eligibleThird-party benchmark
#17Grok 4.20 Multi AgentxAI1,204.80rating—126,238grok-4.20-multi-agent-beta-0309; 95% CI [1199.36322528, 1210.23649279]; votes 126238; rank 18unknownOriginal sourcewinner eligibleThird-party benchmark
#18Claude Opus 4.8Anthropic1,204.11rating—106,304claude-opus-4-8; 95% CI [1198.48669076, 1209.73163091]; votes 106304; rank 19unknownOriginal sourcewinner eligibleThird-party benchmark
#19GPT-5.1OpenAI1,199.70rating—63,042gpt-5.1-search; 95% CI [1194.51255947, 1204.89089087]; votes 63042; rank 20unknownOriginal sourcewinner eligibleThird-party benchmark
#20GPT-5.4OpenAI1,199.08rating—126,730gpt-5.4-search; 95% CI [1193.60193056, 1204.56536227]; votes 126730; rank 21unknownOriginal sourcewinner eligibleThird-party benchmark
#21Gemini 3 FlashGoogle DeepMind1,195.69rating—196,399gemini-3-flash-grounding; 95% CI [1191.39370466, 1199.99583788]; votes 196399; rank 22unknownOriginal sourcewinner eligibleThird-party benchmark
#22Claude Opus 4.5Anthropic1,179.58rating—65,047claude-opus-4-5-search; 95% CI [1173.98777416, 1185.18060388]; votes 65047; rank 25unknownOriginal sourcewinner eligibleThird-party benchmark
#23GPT-6 SolOpenAI1,174.91rating—11,313gpt-6-sol-max; 95% CI [1167.83636087, 1181.98711987]; votes 11313; rank 26unknownOriginal sourcewinner eligibleThird-party benchmark
#24Grok 4.3xAI1,165.81rating—128,393grok-4.3; 95% CI [1161.07351221, 1170.53848666]; votes 128393; rank 30unknownOriginal sourcewinner eligibleThird-party benchmark
#25Claude Sonnet 4.5Anthropic1,160.27rating—144,153claude-sonnet-4-5-search; 95% CI [1155.14751166, 1165.38826265]; votes 144153; rank 31unknownOriginal sourcewinner eligibleThird-party benchmark
#26GPT-6 LunaOpenAI1,159.56rating—10,793gpt-6-luna-max; 95% CI [1152.33057533, 1166.78316335]; votes 10793; rank 32unknownOriginal sourcewinner eligibleThird-party benchmark
#27Gemini 2.5 ProGoogle DeepMind1,142.80rating—86,689gemini-2.5-pro-grounding; 95% CI [1138.12227943, 1147.47015160]; votes 86689; rank 35unknownOriginal sourcewinner eligibleThird-party benchmark
#28Sonar Reasoning ProPerplexity1,139.01rating—29,439ppl-sonar-reasoning-pro-high; 95% CI [1133.19180048, 1144.81939817]; votes 29439; rank 37unknownOriginal sourcewinner eligibleThird-party benchmark
#29GPT-5OpenAI1,133.49rating—20,865gpt-5-search; 95% CI [1127.51131313, 1139.47049228]; votes 20865; rank 38unknownOriginal sourcewinner eligibleThird-party benchmark
#30Sonar ProPerplexitySelected model1,130.28rating—28,899ppl-sonar-pro-high; 95% CI [1124.47147876, 1136.08107434]; votes 28899; rank 39unknownOriginal sourcewinner eligibleThird-party benchmark

30 models ranked by the best winner-eligible current-version score when available, otherwise the best publisher-artifact score; highest first. Observed Oct 19, 2026. The model you came from is highlighted.

Methodology and Coverage

Ranking protocolarena rating · rating. Each model appears once; verified winner-eligible runs take precedence, and missing models are not imputed.
Acquisition coverage30 published models from 42 source rows; 11 source identities remain quarantined.

Primary Evidence

Publisher Artifacts

LMArena Vision Arena methodologyLMArena · observed Oct 19, 2026LMArena Search Arena search-2026-10-13-a64ca3b12b6e score tableLMArena · observed Oct 19, 2026

Questions

LMArena Search Arena FAQs

What does LMArena Search Arena measure?+

Blind human preferences for search-augmented responses in LMArena. Model Markets classifies it as a search benchmark and preserves the publisher's search-2026-10-13-a64ca3b12b6e release as a distinct comparison cohort.

How are models ranked on LMArena Search Arena?+

Models are ordered by arena rating in rating, with higher scores ranked first. Each model appears once; a verified winner-eligible current-version result takes precedence over a publisher-artifact-only result, and missing scores are not estimated.

Which model currently leads LMArena Search Arena?+

GPT-5.6 Sol leads the current verified table with 1,261.13 rating on search-2026-10-13-a64ca3b12b6e. This is a benchmark-specific result, not a universal model-quality claim.

How many models have a published LMArena Search Arena score?+

30 catalog models are published from 42 source rows. 11 source identities remain quarantined rather than guessed.

Can LMArena Search Arena scores be compared with other benchmarks?+

Raw scores should be compared only within the same benchmark version, metric, and protocol. The Intelligence ranking normalizes its three pinned publisher inputs to a frozen reference range, requires at least one source, and labels one- and two-source rows provisional instead of estimating missing inputs. View aggregate rankings →

Why might a model be missing from LMArena Search Arena?+

A model remains absent when the publisher has no current result, the source model identity is unresolved, the evaluation protocol is incompatible, or the evidence cannot be verified. Model Markets does not substitute a provider claim or infer a score from a related model.

How current is the LMArena Search Arena leaderboard?+

The current Model Markets snapshot was observed Oct 19, 2026 from LMArena artifacts. The exact publisher source and each retained result artifact are linked on this page.

Send Feedback