Model Markets
Get Free API KeyAPI

LMArena Text Arena

LMArena · text-2026-10-15-a64ca3b12b6e · Blind pairwise human preferences from LMArena's text arena.

Model Ranking

Source
#1Qwen3.8 MaxQwen1,482.88rating—27,766qwen3.8-max; 95% CI [1477.92809424, 1487.82635626]; votes 27766; rank 17unknownOriginal sourcewinner eligibleThird-party benchmark
#2Gemini 3.5 FlashGoogle DeepMind1,482.02rating—48,486gemini-3.5-flash-high; 95% CI [1478.14855340, 1485.88478693]; votes 48486; rank 18unknownOriginal sourcewinner eligibleThird-party benchmark
#3Gemini 3.6 FlashGoogle DeepMind1,479.31rating—36,122gemini-3.6-flash-high; 95% CI [1474.91726787, 1483.70716010]; votes 36122; rank 21unknownOriginal sourcewinner eligibleThird-party benchmark
#4GLM 5.3Z.ai1,471.46rating—22,543glm-5.3-max; 95% CI [1466.26042381, 1476.65163576]; votes 22543; rank 27unknownOriginal sourcewinner eligibleThird-party benchmark
#5GPT-5.5OpenAI1,471.34rating—69,401gpt-5.5-high; 95% CI [1467.64988983, 1475.02964305]; votes 69401; rank 28unknownOriginal sourcewinner eligibleThird-party benchmark
#6GLM 5.3 FlashZ.ai1,470.53rating—32,925glm-5.3-flash; 95% CI [1465.67476412, 1475.38456118]; votes 32925; rank 29unknownOriginal sourcewinner eligibleThird-party benchmark
#7GLM 5.2Z.ai1,470.37rating—49,820glm-5.2-max; 95% CI [1466.22224854, 1474.51957977]; votes 49820; rank 31unknownOriginal sourcewinner eligibleThird-party benchmark
#8GPT-5.4OpenAI1,468.97rating—64,776gpt-5.4-high; 95% CI [1465.23802304, 1472.70880445]; votes 64776; rank 33unknownOriginal sourcewinner eligibleThird-party benchmark
#9ERNIE 5.1Baidu1,467.73rating—39,468ernie-5.1; 95% CI [1463.22136313, 1472.23707800]; votes 39468; rank 34unknownOriginal sourcewinner eligibleThird-party benchmark
#10Gemini 3 FlashGoogle DeepMind1,465.71rating—31,325gemini-3-flash; 95% CI [1461.39670282, 1470.02720080]; votes 31325; rank 36unknownOriginal sourcewinner eligibleThird-party benchmark
#11GLM 5.1Z.ai1,462.40rating—62,889glm-5.1; 95% CI [1458.83486169, 1465.97294166]; votes 62889; rank 39unknownOriginal sourcewinner eligibleThird-party benchmark
#12Claude Opus 4.8Anthropic1,461.08rating—64,100claude-opus-4-8-high; 95% CI [1457.20719685, 1464.94408250]; votes 64100; rank 40unknownOriginal sourcewinner eligibleThird-party benchmark
#13Claude Sonnet 4.6Anthropic1,457.87rating—70,931claude-sonnet-4-6; 95% CI [1454.37156748, 1461.36302648]; votes 70931; rank 41unknownOriginal sourcewinner eligibleThird-party benchmark
#14Gemini 2.5 ProGoogle DeepMind1,457.73rating—125,366gemini-2.5-pro; 95% CI [1455.29194111, 1460.15968303]; votes 125366; rank 42unknownOriginal sourcewinner eligibleThird-party benchmark
#15GPT-5.6 SolOpenAI1,457.06rating—40,442gpt-5.6-sol-xhigh; 95% CI [1452.73485432, 1461.39117296]; votes 40442; rank 43unknownOriginal sourcewinner eligibleThird-party benchmark
#16Kimi K2.6Moonshot AI1,455.42rating—40,013kimi-k2.6; 95% CI [1451.01237282, 1459.83173802]; votes 40013; rank 44unknownOriginal sourcewinner eligibleThird-party benchmark
#17Qwen3.7 PlusQwen1,454.58rating—42,075qwen3.7-plus; 95% CI [1450.26283424, 1458.90144791]; votes 42075; rank 46unknownOriginal sourcewinner eligibleThird-party benchmark
#18DeepSeek V4 ProDeepSeek1,451.16rating—57,594deepseek-v4-pro; 95% CI [1447.27741116, 1455.03943137]; votes 57594; rank 49unknownOriginal sourcewinner eligibleThird-party benchmark
#19Claude Opus 4.5Anthropic1,450.88rating—72,899claude-opus-4-5-20251101; 95% CI [1447.77007872, 1453.98289133]; votes 72899; rank 50unknownOriginal sourcewinner eligibleThird-party benchmark
#20Grok 4.20 Multi AgentxAI1,450.27rating—64,946grok-4.20-multi-agent-beta-0309; 95% CI [1446.59423552, 1453.95404223]; votes 64946; rank 52unknownOriginal sourcewinner eligibleThird-party benchmark
#21Grok 4.5xAI1,447.97rating—39,896grok-4.5; 95% CI [1443.63647354, 1452.30434473]; votes 39896; rank 56unknownOriginal sourcewinner eligibleThird-party benchmark
#22GPT-6.1 SolOpenAI1,447.67rating—6,320gpt-6.1-sol-max; 95% CI [1439.59979847, 1455.74908604]; votes 6320; rank 57unknownOriginal sourcewinner eligibleThird-party benchmark
#23GPT-5.6 TerraOpenAI1,447.39rating—43,271gpt-5.6-terra-xhigh; 95% CI [1443.15823569, 1451.62460535]; votes 43271; rank 58unknownOriginal sourcewinner eligibleThird-party benchmark
#24GLM 5Z.ai1,445.93rating—29,272glm-5; 95% CI [1441.71120059, 1450.14553038]; votes 29272; rank 60unknownOriginal sourcewinner eligibleThird-party benchmark
#25DeepSeek V4 Pro 0813DeepSeekSelected model1,445.55rating—9,800deepseek-v4-pro-high-20260813; 95% CI [1438.98161909, 1452.11869256]; votes 9800; rank 61unknownOriginal sourcewinner eligibleThird-party benchmark
#26Kimi K2.5Moonshot AI1,445.02rating—74,759kimi-k2.5-thinking; 95% CI [1441.75782956, 1448.27322629]; votes 74759; rank 62unknownOriginal sourcewinner eligibleThird-party benchmark
#27Gemma 4 31BGoogle DeepMind1,442.86rating—6,147gemma-4-31b; 95% CI [1435.47112822, 1450.25595360]; votes 6147; rank 67unknownOriginal sourcewinner eligibleThird-party benchmark
#28GPT-5.1OpenAI1,442.74rating—41,651gpt-5.1-high; 95% CI [1439.06360530, 1446.41090056]; votes 41651; rank 68unknownOriginal sourcewinner eligibleThird-party benchmark
#29GPT-6 AstraOpenAI1,442.69rating—12,163gpt-6-astra-max; 95% CI [1436.40189535, 1448.98654490]; votes 12163; rank 69unknownOriginal sourcewinner eligibleThird-party benchmark
#30Claude Sonnet 5Anthropic1,442.12rating—49,595claude-sonnet-5-high; 95% CI [1438.01344792, 1446.22951308]; votes 49595; rank 72unknownOriginal sourcewinner eligibleThird-party benchmark
#31InklingThinking Machines Lab1,441.38rating—40,648inkling; 95% CI [1437.08506747, 1445.68368363]; votes 40648; rank 74unknownOriginal sourcewinner eligibleThird-party benchmark
#32Qwen3.8 27BQwen1,440.55rating—23,294qwen3.8-27b; 95% CI [1435.47949951, 1445.61514077]; votes 23294; rank 75unknownOriginal sourcewinner eligibleThird-party benchmark
#33Claude Sonnet 4.5Anthropic1,438.42rating—82,688claude-sonnet-4-5-20250929; 95% CI [1435.57332334, 1441.27604162]; votes 82688; rank 79unknownOriginal sourcewinner eligibleThird-party benchmark
#34Hy3Tencent1,438.00rating—12,208hy3; 95% CI [1431.86907820, 1444.13361939]; votes 12208; rank 80unknownOriginal sourcewinner eligibleThird-party benchmark
#35Qwen3.5 397B A17BQwen1,437.72rating—93,567qwen3.5-397b-a17b; 95% CI [1434.65086689, 1440.78699055]; votes 93567; rank 81unknownOriginal sourcewinner eligibleThird-party benchmark
#36Gemini 3.5 Flash LiteGoogle DeepMind1,434.52rating—35,851gemini-3.5-flash-lite; 95% CI [1430.05687105, 1438.97565121]; votes 35851; rank 89unknownOriginal sourcewinner eligibleThird-party benchmark
#37Gemma 4 26B A4BGoogle DeepMind1,434.12rating—6,094gemma-4-26b-a4b; 95% CI [1426.70039347, 1441.53160125]; votes 6094; rank 90unknownOriginal sourcewinner eligibleThird-party benchmark
#38MiniMax M3MiniMax1,433.33rating—63,973minimax-m3; 95% CI [1429.51159020, 1437.14164071]; votes 63973; rank 91unknownOriginal sourcewinner eligibleThird-party benchmark
#39DeepSeek V4 FlashDeepSeek1,431.84rating—51,987deepseek-v4-flash; 95% CI [1427.81979734, 1435.85552908]; votes 51987; rank 92unknownOriginal sourcewinner eligibleThird-party benchmark
#40GPT-5.6 LunaOpenAI1,430.66rating—41,109gpt-5.6-luna-xhigh; 95% CI [1426.38028671, 1434.93119450]; votes 41109; rank 93unknownOriginal sourcewinner eligibleThird-party benchmark
#41Mistral Large 3Mistral AI1,428.16rating—84,205mistral-large-3; 95% CI [1425.33671505, 1430.99053174]; votes 84205; rank 98unknownOriginal sourcewinner eligibleThird-party benchmark
#42Grok 4.6xAI1,427.77rating—28,965grok-4.6-high; 95% CI [1422.91144511, 1432.61896218]; votes 28965; rank 99unknownOriginal sourcewinner eligibleThird-party benchmark
#43DeepSeek V3.2DeepSeek1,424.46rating—48,112deepseek-v3.2; 95% CI [1420.99052972, 1427.92299771]; votes 48112; rank 105unknownOriginal sourcewinner eligibleThird-party benchmark
#44Mistral Medium 3.5 128BMistral AI1,420.97rating—11,778mistral-medium-3.5; 95% CI [1414.59009121, 1427.34230551]; votes 11778; rank 112unknownOriginal sourcewinner eligibleThird-party benchmark
#45Gemini 2.5 FlashGoogle DeepMind1,416.89rating—125,521gemini-2.5-flash; 95% CI [1414.48804539, 1419.29483183]; votes 125521; rank 127unknownOriginal sourcewinner eligibleThird-party benchmark
#46GPT-5.2OpenAI1,416.24rating—49,655gpt-5.2-high; 95% CI [1412.72461401, 1419.75769766]; votes 49655; rank 129unknownOriginal sourcewinner eligibleThird-party benchmark
#47Inkling SmallThinking Machines Lab1,412.62rating—32,304Inkling Small; 95% CI [1407.97725962, 1417.26957820]; votes 32304; rank 135unknownOriginal sourcewinner eligibleThird-party benchmark
#48GPT-5.4 MiniOpenAI1,411.45rating—63,449gpt-5.4-mini-high; 95% CI [1407.73358875, 1415.16901218]; votes 63449; rank 137unknownOriginal sourcewinner eligibleThird-party benchmark
#49GPT-5OpenAI1,406.18rating—31,722gpt-5-high; 95% CI [1401.76102728, 1410.60028213]; votes 31722; rank 146unknownOriginal sourcewinner eligibleThird-party benchmark
#50MiniMax M2.7MiniMax1,404.83rating—84,977minimax-m2.7; 95% CI [1401.50702597, 1408.15721867]; votes 84977; rank 148unknownOriginal sourcewinner eligibleThird-party benchmark
#51GPT-6 SolOpenAI1,398.02rating—12,466gpt-6-sol-max; 95% CI [1391.69518333, 1404.34476736]; votes 12466; rank 156unknownOriginal sourcewinner eligibleThird-party benchmark
#52Grok 4.3xAI1,396.92rating—71,347grok-4.3; 95% CI [1393.31445966, 1400.53298228]; votes 71347; rank 158unknownOriginal sourcewinner eligibleThird-party benchmark
#53Claude Haiku 4.5Anthropic1,396.64rating—149,199claude-haiku-4-5-20251001; 95% CI [1394.24930739, 1399.03971640]; votes 149199; rank 159unknownOriginal sourcewinner eligibleThird-party benchmark
#54Qwen3.5 35B A3BQwen1,394.73rating—30,457qwen3.5-35b-a3b; 95% CI [1390.52780278, 1398.93686884]; votes 30457; rank 161unknownOriginal sourcewinner eligibleThird-party benchmark
#55GPT-6 LunaOpenAI1,391.63rating—49,617gpt-6-luna-max; 95% CI [1387.01005073, 1396.24774606]; votes 49617; rank 166unknownOriginal sourcewinner eligibleThird-party benchmark
#56Muse Glimmer 30BMeta1,388.91rating—3,897muse-glimmer; 95% CI [1379.26475817, 1398.55311930]; votes 3897; rank 167unknownOriginal sourcewinner eligibleThird-party benchmark
#57Solar Pro 4Upstage1,386.72rating—21,281solar-pro4; 95% CI [1381.43817825, 1392.00326711]; votes 21281; rank 169unknownOriginal sourcewinner eligibleThird-party benchmark
#58GPT-4.1OpenAI1,382.89rating—50,312gpt-4.1-2025-04-14; 95% CI [1379.19934172, 1386.58437112]; votes 50312; rank 172unknownOriginal sourcewinner eligibleThird-party benchmark
#59GPT-5 MiniOpenAI1,373.28rating—26,804gpt-5-mini-high; 95% CI [1368.72762571, 1377.83899204]; votes 26804; rank 180unknownOriginal sourcewinner eligibleThird-party benchmark
#60DeepSeek R1DeepSeek1,372.65rating—18,524deepseek-r1; 95% CI [1367.80631687, 1377.49940563]; votes 18524; rank 181unknownOriginal sourcewinner eligibleThird-party benchmark
#61GPT-5.4 NanoOpenAI1,372.01rating—62,396gpt-5.4-nano-high; 95% CI [1368.25704083, 1375.76986498]; votes 62396; rank 182unknownOriginal sourcewinner eligibleThird-party benchmark
#62gpt-oss-120bOpenAI1,365.51rating—30,206gpt-oss-120b; 95% CI [1361.16495776, 1369.85717870]; votes 30206; rank 192unknownOriginal sourcewinner eligibleThird-party benchmark
#63Nova 2 LiteAmazon1,361.83rating—12,413nova-2-lite; 95% CI [1355.81915120, 1367.83207069]; votes 12413; rank 195unknownOriginal sourcewinner eligibleThird-party benchmark
#64Granite 4.2 30BIBM1,361.41rating—3,401granite-4.2-30b; 95% CI [1350.97772869, 1371.83470317]; votes 3401; rank 196unknownOriginal sourcewinner eligibleThird-party benchmark
#65MiniMax M2.5MiniMax1,359.12rating—42,815minimax-m2.5; 95% CI [1355.25588729, 1362.99197980]; votes 42815; rank 198unknownOriginal sourcewinner eligibleThird-party benchmark
#66NVIDIA Nemotron 3 Nano 30B A3B BF16NVIDIA1,347.82rating—15,770nvidia-nemotron-3-nano-30b-a3b-bf16; 95% CI [1342.43511795, 1353.20837526]; votes 15770; rank 209unknownOriginal sourcewinner eligibleThird-party benchmark
#67GPT-4.1 MiniOpenAI1,340.50rating—38,709gpt-4.1-mini-2025-04-14; 95% CI [1336.20142291, 1344.80800339]; votes 38709; rank 211unknownOriginal sourcewinner eligibleThird-party benchmark
#68DeepSeek V3DeepSeek1,332.46rating—21,770deepseek-v3; 95% CI [1327.72698428, 1337.19353550]; votes 21770; rank 224unknownOriginal sourcewinner eligibleThird-party benchmark
#69Command ACohere1,330.94rating—55,766command-a-03-2025; 95% CI [1327.49064818, 1334.38717569]; votes 55766; rank 225unknownOriginal sourcewinner eligibleThird-party benchmark
#70NVIDIA Nemotron 3.5 Lightning 30B A3BNVIDIA1,328.88rating—19,938nvidia-nemotron-3.5-lightning-30b-a3b-nvfp4; 95% CI [1323.35291228, 1334.41299389]; votes 19938; rank 229unknownOriginal sourcewinner eligibleThird-party benchmark
#71GPT-5 NanoOpenAI1,319.94rating—8,132gpt-5-nano-high; 95% CI [1313.09185399, 1326.78218151]; votes 8132; rank 232unknownOriginal sourcewinner eligibleThird-party benchmark
#72Granite 4.2 8BIBM1,319.58rating—3,254granite-4.2-8b; 95% CI [1308.51127780, 1330.64584507]; votes 3254; rank 234unknownOriginal sourcewinner eligibleThird-party benchmark
#73Olmo 3.1 32B InstructAi21,311.14rating—11,817olmo-3.1-32b-instruct; 95% CI [1305.04954142, 1317.22764633]; votes 11817; rank 242unknownOriginal sourcewinner eligibleThird-party benchmark
#74Olmo 3 32B ThinkAi21,299.42rating—5,831olmo-3-32b-think; 95% CI [1291.17882644, 1307.65677276]; votes 5831; rank 248unknownOriginal sourcewinner eligibleThird-party benchmark
#75Granite 4.2 3BIBM1,292.94rating—3,251granite-4.2-3b; 95% CI [1281.43969143, 1304.43063080]; votes 3251; rank 253unknownOriginal sourcewinner eligibleThird-party benchmark
#76Granite 4.1 8BIBM1,290.69rating—4,180granite-4.1-8b; 95% CI [1280.76464094, 1300.60785998]; votes 4180; rank 256unknownOriginal sourcewinner eligibleThird-party benchmark
#77gpt-oss-20bOpenAI1,287.43rating—10,419gpt-oss-20b; 95% CI [1281.03637900, 1293.81371199]; votes 10419; rank 260unknownOriginal sourcewinner eligibleThird-party benchmark
#78Llama 4 Maverick 17B 128E InstructMeta1,287.19rating—39,450llama-4-maverick-17b-128e-instruct; 95% CI [1282.93484207, 1291.44351277]; votes 39450; rank 261unknownOriginal sourcewinner eligibleThird-party benchmark
#79GPT-4o MiniOpenAI1,286.34rating—68,705gpt-4o-mini-2024-07-18; 95% CI [1282.82219462, 1289.86344250]; votes 68705; rank 263unknownOriginal sourcewinner eligibleThird-party benchmark
#80GPT-4.1 NanoOpenAI1,284.78rating—6,103gpt-4.1-nano-2025-04-14; 95% CI [1277.13551573, 1292.41837024]; votes 6103; rank 264unknownOriginal sourcewinner eligibleThird-party benchmark
#81GPT-4oOpenAI1,282.50rating—45,499gpt-4o-2024-08-06; 95% CI [1278.31275237, 1286.69364786]; votes 45499; rank 267unknownOriginal sourcewinner eligibleThird-party benchmark
#82Llama 4 Scout 17B 16E InstructMeta1,279.16rating—29,826llama-4-scout-17b-16e-instruct; 95% CI [1274.44779094, 1283.86814615]; votes 29826; rank 273unknownOriginal sourcewinner eligibleThird-party benchmark
#83Llama 3.3 70B InstructMeta1,274.21rating—54,497llama-3.3-70b-instruct; 95% CI [1270.72459081, 1277.68627453]; votes 54497; rank 276unknownOriginal sourcewinner eligibleThird-party benchmark
#84Llama 3.1 70B InstructMeta1,260.87rating—55,240llama-3.1-70b-instruct; 95% CI [1257.15696595, 1264.58148679]; votes 55240; rank 290unknownOriginal sourcewinner eligibleThird-party benchmark
#85phi-4Microsoft1,216.66rating—24,126phi-4; 95% CI [1212.07369450, 1221.23689100]; votes 24126; rank 314unknownOriginal sourcewinner eligibleThird-party benchmark
#86Llama 3.1 8B InstructMeta1,186.54rating—49,605llama-3.1-8b-instruct; 95% CI [1182.43982822, 1190.63886318]; votes 49605; rank 327unknownOriginal sourcewinner eligibleThird-party benchmark

86 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 coverage86 published models from 416 source rows; 321 source identities remain quarantined.

Primary Evidence

Publisher Artifacts

LMArena Vision Arena methodologyLMArena · observed Oct 19, 2026LMArena Text Arena text-2026-10-15-a64ca3b12b6e score tableLMArena · observed Oct 19, 2026

Questions

LMArena Text Arena FAQs

What does LMArena Text Arena measure?+

Blind pairwise human preferences from LMArena's text arena. Model Markets classifies it as a general language benchmark and preserves the publisher's text-2026-10-15-a64ca3b12b6e release as a distinct comparison cohort.

How are models ranked on LMArena Text 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 Text Arena?+

Qwen3.8 Max leads the current verified table with 1,482.88 rating on text-2026-10-15-a64ca3b12b6e. This is a benchmark-specific result, not a universal model-quality claim.

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

86 catalog models are published from 416 source rows. 321 source identities remain quarantined rather than guessed.

Can LMArena Text 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 Text 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 Text 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