TestsBenchmarks · Human preference
LMArena Text - Multi-Turn
Crowd-voted, style-controlled Elo on multi-turn conversations (iterative drafting/revision).
Elo ratingHigher is betterCounts toward:Weights: Writing & creativity ×0.5The test's websiteOfficial page ↗
Measured by independent testersIndependent Reported by the makerVendor-reported
| Thinking levelSetting | |||||||
|---|---|---|---|---|---|---|---|
| 1 | Gemini 4 Argon | Google (Gemini / DeepMind) | 1552 | Highgemini-4-argon-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 2 | Muse Spark 1.2 | Meta (Meta Superintelligence Labs, Muse) | 1522 | Extra highmuse-spark-1.2 (xHigh) | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 3 | Claude Opus 4.6 | Anthropic | 1519 | Highclaude-opus-4-6-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 4 | Claude Fable 5 | Anthropic | 1518 | Highclaude-fable-5-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 5 | Claude Opus 4.7 | Anthropic | 1516 | Highclaude-opus-4-7-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 6 | Gemini 3.8 Flash | Google (Gemini / DeepMind) | 1501 | Highgemini-3.8-flash-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 7 | Kimi K3 | Moonshot AI (Kimi) | 1500 | Maxkimi-k3-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 8 | Claude Opus 4.8 | Anthropic | 1498 | Highclaude-opus-4-8-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 9 | Claude Opus 5.5 | Anthropic | 1497 | Highclaude-opus-5.5-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 10 | Gemini 3.1 Pro (Preview) | Google (Gemini / DeepMind) | 1497 | Defaultgemini-3.1-pro-preview | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 11 | Muse Spark 1.1 | Meta (Meta Superintelligence Labs, Muse) | 1496 | Defaultmuse-spark-1.1 | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 12 | Gemini 3 Pro | Google (Gemini / DeepMind) | 1496 | Defaultgemini-3-pro | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 13 | Gemini 3.7 Flash | Google (Gemini / DeepMind) | 1496 | Highgemini-3.7-flash-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 14 | GPT-5.2 Chat (ChatGPT snapshot) | OpenAI | 1495 | Defaultgpt-5.2-chat-latest-20260210 | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 15 | Muse Spark | Meta (Meta Superintelligence Labs, Muse) | 1494 | Defaultmuse-spark | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 16 | GPT-5.4 | OpenAI | 1493 | Highgpt-5.4-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 17 | Muse Spark 1.3 | Meta (Meta Superintelligence Labs, Muse) | 1491 | Maxmuse-spark-1.3-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 18 | Qwen3.8 Max (0803) | Qwen (Alibaba) | 1490 | Defaultqwen3.8-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 19 | GPT-6 Astra | OpenAI | 1489 | Maxgpt-6-astra-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 20 | Claude Fable 5.1 | Anthropic | 1487 | Maxclaude-fable-5.1-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 21 | GLM-5.3 | Z.ai (Zhipu) | 1481 | Maxglm-5.3-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 22 | Claude Sonnet 5 | Anthropic | 1474 | Highclaude-sonnet-5-high | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 23 | DeepSeek V4 Pro (0813) | DeepSeek | 1472 | Highdeepseek-v4-pro-high-20260813 | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 24 | GPT-6 Sol | OpenAI | 1469 | Maxgpt-6-sol-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 25 | DeepSeek V4.1 Flash | DeepSeek | 1462 | Maxdeepseek-v4.1-flash-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 26 | MiMo-V2.6-Pro | Xiaomi MiMo | 1455 | Defaultmimo-v2.6-pro | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 27 | GPT-6 Luna | OpenAI | 1455 | Maxgpt-6-luna-max | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |
| 28 | Grok 4.7 | SpaceXAI (formerly xAI) | 1425 | Extra highgrok-4.7-xhigh | Independent testIndependentLMArena ↗ | — | 30 Sep 2026 |