LMArena Longer Queries
Text arena rating on long prompts.
As of 19 Sept 2026, Claude Opus 4.6 leads LMArena Longer Queries on BenchLeader with 1524, ahead of Claude Fable 5 at 1521, across 344 model configurations with a published result.
- Published by
- LMArena
- Category
- Human preference
- Index weight
- Reference only
- Models
- 344
- Data as of
- 19 Sept 2026
CC BY 4.0 — lmarena-ai/leaderboard-dataset on Hugging Face.
What the test looks like
Pairwise votes on prompts above a length threshold.
How it is scored
Bradley-Terry rating from pairwise votes, in Elo-like units, with style control.
What to keep in mind
Fewer votes than the overall board.
- 1Claude Opus 4.6 (high)1524
- 2Claude Fable 51521
- 3Claude Opus 4.61517
- 4Claude Opus 4.7 (high)1514
- 5Claude Fable 5.1 (max)1511
- 6Gemini 3.8 Flash (high)1509
- 7Claude Opus 5 (high)1508
- 8Claude Opus 4.71507
- 9Muse Spark 1.3 (max)1506
- 10Muse Spark 1.2 (xhigh)1502
- 11Claude Opus 4.8 (high)1501
- 12Kimi K3 (max)1500
- 13Gemini 3.1 Pro1500
- 14Claude Opus 4.81500
- 15Claude Opus 5 (max)1500
344 of 344
| # | ||||
|---|---|---|---|---|
| 1 |  Claude Opus 4.6highAnthropic | 1524 | 61.3 | 2026-09-13 |
| 2 | Claude Fable 5Anthropic | 1521 | 68.3 | 2026-09-13 |
| 3 | Claude Opus 4.6Anthropic | 1517 | 63.7 | 2026-09-13 |
| 4 | Claude Opus 4.7highAnthropic | 1514 | 62.4 | 2026-09-13 |
| 5 | Claude Fable 5.1maxAnthropic | 1511 | 69.7 | 2026-09-13 |
| 6 |  Gemini 3.8 FlashhighGoogle | 1509 | 64.5 | 2026-09-13 |
| 7 | Claude Opus 5highAnthropic | 1508 | 70.2 | 2026-09-13 |
| 8 | Claude Opus 4.7Anthropic | 1507 | 64.5 | 2026-09-13 |
| 9 |  Muse Spark 1.3maxMeta | 1506 | 69.3 | 2026-09-13 |
| 10 | Muse Spark 1.2xhighMeta | 1502 | 64.1 | 2026-09-13 |
| 11 | Claude Opus 4.8highAnthropic | 1501 | 62.2 | 2026-09-13 |
| 12 |  Kimi K3maxMoonshot AIopen ↗ | 1500 | 67.2 | 2026-09-13 |
| 13 | Gemini 3.1 ProGoogle | 1500 | 63.9 | 2026-09-13 |
| 14 | Claude Opus 4.8Anthropic | 1500 | 61.9 | 2026-09-13 |
| 15 | Claude Opus 5maxAnthropic | 1500 | 69.9 | 2026-09-13 |
| 16 | Gemini 3.7 FlashhighGoogle | 1498 | 64.6 | 2026-09-13 |
| 17 |  GPT-5.6 SolxhighOpenAI | 1498 | 68.2 | 2026-09-13 |
| 18 | Claude Opus 4.5highAnthropic | 1494 | 57.7 | 2026-09-13 |
| 19 | Claude Sonnet 4.6Anthropic | 1494 | 59.0 | 2026-09-13 |
| 20 |  Qwen3 7maxAlibaba | 1494 | 61.0 | 2026-09-13 |
| 21 | Claude Opus 4.5Anthropic | 1493 | 58.2 | 2026-09-13 |
| 22 |  GLM-5.3maxZhipu AIopen ↗ | 1492 | 65.8 | 2026-09-13 |
| 23 | Qwen3 8maxAlibaba | 1492 | 66.5 | 2026-09-13 |
| 24 | Gemini 3 ProGoogle | 1491 | 61.1 | 2026-09-13 |
| 25 | GPT-5.5highOpenAI | 1488 | 67.0 | 2026-09-13 |
| 26 | Gemini 3.6 FlashhighGoogle | 1487 | 61.7 | 2026-09-13 |
| 27 |  MiMo-V2.5-ProXiaomiopen ↗ | 1487 | 59.2 | 2026-09-13 |
| 28 | Claude Sonnet 4.5highAnthropic | 1486 | 58.8 | 2026-09-13 |
| 29 | GPT-5.5OpenAI | 1485 | 63.2 | 2026-09-13 |
| 30 | Claude Opus 4.1thinkingAnthropic | 1485 | 57.2 | 2026-09-13 |
| 31 | GLM-5.2maxZhipu AIopen ↗ | 1483 | 63.9 | 2026-09-13 |
| 32 | GPT-5.4highOpenAI | 1483 | 59.0 | 2026-09-13 |
| 33 |  Grok 4.5xAI | 1483 | 60.3 | 2026-09-13 |
| 34 | Claude Sonnet 4.5Anthropic | 1483 | 54.3 | 2026-09-13 |
| 35 | Claude Sonnet 5highAnthropic | 1482 | 62.2 | 2026-09-13 |
| 36 | Qwen3 5maxAlibaba | 1482 | – | 2026-09-13 |
| 37 | GPT-6 AstramaxOpenAI | 1481 | 71.8 | 2026-09-13 |
| 38 | Gemini 3.5 FlashhighGoogle | 1481 | 63.6 | 2026-09-13 |
| 39 | Muse Spark 1.1Meta | 1481 | 65.2 | 2026-09-13 |
| 40 | GLM-5.1Zhipu AIopen ↗ | 1480 | 56.9 | 2026-09-13 |
| 41 | GPT-5.4OpenAI | 1478 | 59.2 | 2026-09-13 |
| 42 | Gemini 3.5 FlashmediumGoogle | 1478 | 64.2 | 2026-09-13 |
| 43 | Gemini 3 FlashGoogle | 1477 | 57.7 | 2026-09-13 |
| 44 | GLM-5.3-FlashZhipu AIopen ↗ | 1476 | 63.9 | 2026-09-13 |
| 45 | Kimi K2.6Moonshot AIopen ↗ | 1476 | 60.5 | 2026-09-13 |
| 46 | GPT-5.5 InstantOpenAI | 1475 | 59.2 | 2026-09-13 |
| 47 | Muse SparkMeta | 1474 | 65.7 | 2026-09-13 |
| 48 | Qwen3 6maxAlibaba | 1474 | 63.3 | 2026-09-13 |
| 49 |  Deepseek v4 ProhighDeepSeek | 1474 | 60.8 | 2026-09-13 |
| 50 | GPT-5.2OpenAI | 1474 | 58.4 | 2026-09-13 |
| 51 | Claude Opus 4.1Anthropic | 1472 | 51.9 | 2026-09-13 |
| 52 | DeepSeek V4 ProDeepSeekopen ↗ | 1472 | 52.6 | 2026-09-13 |
| 53 | Grok 4.6highxAI | 1471 | 64.5 | 2026-09-13 |
| 54 | GLM-5Zhipu AIopen | 1470 | 54.7 | 2026-09-13 |
| 55 | GPT-5.6 TerraxhighOpenAI | 1468 | 64.3 | 2026-09-13 |
| 56 | Gemma 4 31BGoogleopen ↗ | 1468 | 52.7 | 2026-09-13 |
| 57 | Mimo v2 ProXiaomi | 1467 | 59.8 | 2026-09-13 |
| 58 |  Ernie 5.1Baidu | 1466 | – | 2026-09-13 |
| 59 | Claude Opus 4thinkingAnthropic | 1466 | 55.7 | 2026-09-13 |
| 60 | Grok 4.20thinkingxAI | 1465 | 59.9 | 2026-09-13 |
Cite as: BenchLeader, “LMArena Longer Queries leaderboard”, https://www.benchleader.com/benchmarks/lmarena_text_longer, data as of 19 Sept 2026.
LMArena Longer Queries: questions
- What does LMArena Longer Queries measure?
- Pairwise votes on prompts above a length threshold. Scores are reported in Elo-style ratings from pairwise comparisons; higher is better.
- Which AI model leads LMArena Longer Queries?
- Claude Opus 4.6 leads LMArena Longer Queries with 1524 as of 19 Sept 2026, ahead of Claude Fable 5 at 1521.
- How many models have LMArena Longer Queries results?
- 344 model configurations have a LMArena Longer Queries result on BenchLeader, all taken from LMArena.
- Who runs LMArena Longer Queries and how often is it updated?
- LMArena Longer Queries is published by LMArena. BenchLeader re-reads the published results every morning and records the date each result was published.
- Does LMArena Longer Queries count toward the BenchLeader Index?
- No. LMArena Longer Queries is shown for reference but left out of the composite index.