Benchmark leaderboard - 50 entries on 2 boards
LMArena leaderboard
LMArena ranks AI models on two boards: Text, covering math, coding, creative writing and other open-ended tasks, and WebDev, covering front-end web development including multi-step agentic coding with tool use. As of October 17, 2026, Gemini 4 Argon (High) (Google) ranks first on the LMArena Text (Text Arena Overall) board with 1524±9 (Preliminary) and Claude Opus 5.5 (Max) (Anthropic) ranks first on the LMArena WebDev (Code Arena WebDev Overall) board with 1807+14/-14.
Last verified from arena.ai. Read the caveats
This page covers 2 separate boards: LMArena Text (Text Arena Overall) and LMArena WebDev (Code Arena WebDev Overall). Each is scored on its own terms, so compare rows only within the same board.
LMArena Text (Text Arena Overall)
25 entries, in the board's order
| 1 | Gemini 4 Argon (High) (Google) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 1 (rank spread 1-3); 4,904 votes; license: Proprietary Official leaderboard | 1524±9 (Preliminary) |
| 2 | Muse Wonder (Max) (Meta) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 2 (rank spread 1-11); 3,247 votes; license: Proprietary Official leaderboard | 1509±11 |
| 3 | Claude Opus 5.5 (High) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 3 (rank spread 1-8); 8,300 votes; license: Proprietary Official leaderboard | 1508±7 |
| 4 | Claude Fable 5 (High) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 4 (rank spread 2-11); 42,004 votes; license: Proprietary Official leaderboard | 1504±4 |
| 5 | Claude Opus 4.6 (High) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 5 (rank spread 2-11); 81,394 votes; license: Proprietary Official leaderboard | 1503±3 |
| 6 | Claude Fable 5.1 (Max) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 6 (rank spread 2-12); 14,590 votes; license: Proprietary Official leaderboard | 1502±6 |
| 7 | Claude Opus 4.7 (High) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 7 (rank spread 2-12); 65,116 votes; license: Proprietary Official leaderboard | 1501±4 |
| 8 | Claude Opus 4.6 (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 8 (rank spread 3-15); 86,071 votes; license: Proprietary Official leaderboard | 1497±3 |
| 9 | Gemini 3.8 Flash (High) (Google) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 9 (rank spread 2-18); 37,454 votes; license: Proprietary Official leaderboard | 1496±5 (Preliminary) |
| 10 | Muse Spark 1.3 (Max) (Meta) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 10 (rank spread 3-22); 17,561 votes; license: Proprietary Official leaderboard | 1494±6 |
| 11 | Claude Opus 4.7 (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 11 (rank spread 6-20); 66,216 votes; license: Proprietary Official leaderboard | 1494±4 |
| 12 | Muse Spark 1.2 (xHigh) (Meta) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 12 (rank spread 3-25); 8,248 votes; license: Proprietary Official leaderboard | 1493±7 |
| 13 | Muse Spark 1.1 (Meta) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 13 (rank spread 8-25); 42,810 votes; license: Proprietary Official leaderboard | 1490±4 |
| 14 | Claude Opus 5 (Max) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 14 (rank spread 8-27); 32,495 votes; license: Proprietary Official leaderboard | 1490±5 |
| 15 | Claude Opus 5 (High) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 15 (rank spread 9-25); 66,824 votes; license: Proprietary Official leaderboard | 1490±4 |
| 16 | Muse Spark (Meta) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 16 (rank spread 8-30); 14,150 votes; license: Proprietary Official leaderboard | 1489±6 |
| 17 | Gemini 3.7 Flash (High) (Google) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 17 (rank spread 9-30); 27,232 votes; license: Proprietary Official leaderboard | 1488±5 (Preliminary) |
| 18 | Kimi K3 (Max) (Moonshot) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 18 (rank spread 10-31); 32,795 votes; license: Kimi K3 license Official leaderboard | 1487±5 |
| 19 | Gemini 3.1 Pro Preview (Google) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 19 (rank spread 10-28); 127,649 votes; license: Proprietary Official leaderboard | 1487±3 |
| 20 | Gemini 3 Pro (Google) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 20 (rank spread 11-31); 42,072 votes; license: Proprietary Official leaderboard | 1485±4 |
| 21 | GPT-5.6 Sol (xHigh) (OpenAI) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 21 (rank spread 11-33); 40,442 votes; license: Proprietary Official leaderboard | 1485±4 |
| 22 | GPT-6.1 Sol (Max) (OpenAI) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 22 (rank spread 9-44); 6,320 votes; license: Proprietary Official leaderboard | 1484±8 |
| 23 | Gemini 3.6 Flash (High) (Google) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 23 (rank spread 12-40); 36,122 votes; license: Proprietary Official leaderboard | 1483±4 |
| 24 | Qwen3.8-Max (Alibaba) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 24 (rank spread 12-43); 27,766 votes; license: Proprietary Official leaderboard | 1482±5 |
| 25 | Claude Opus 4.8 (High) (Anthropic) Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 25 (rank spread 15-42); 64,100 votes; license: Proprietary Official leaderboard | 1481±4 |
LMArena WebDev (Code Arena WebDev Overall)
25 entries, in the board's order
| 1 | Claude Opus 5.5 (Max) (Anthropic) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 1 (rank spread 1-1); 2,775 votes; license: Proprietary Official leaderboard | 1807+14/-14 |
| 2 | GPT-6 Astra (Max) (OpenAI) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 2 (rank spread 2-4); 6,014 votes; license: Proprietary Official leaderboard | 1777+10/-10 |
| 3 | Claude Sonnet 5.5 (xHigh) (Anthropic) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 3 (rank spread 2-4); 3,182 votes; license: Proprietary Official leaderboard | 1768+13/-13 |
| 4 | GPT-6.1 Sol (Max) (OpenAI) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 4 (rank spread 2-5); 2,863 votes; license: Proprietary Official leaderboard | 1755+13/-13 |
| 5 | Claude Fable 5.1 (Max) (Anthropic) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 5 (rank spread 4-5); 7,503 votes; license: Proprietary Official leaderboard | 1741+9/-9 |
| 6 | Claude Sonnet 5.5 (High) (Anthropic) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 6 (rank spread 6-6); 4,914 votes; license: Proprietary Official leaderboard | 1711+10/-10 |
| 7 | Claude Opus 5 (Max) (Anthropic) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 7 (rank spread 7-10); 18,310 votes; license: Proprietary Official leaderboard | 1689+6/-6 |
| 8 | GPT-6 Sol (Max) (OpenAI) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 8 (rank spread 7-11); 4,364 votes; license: Proprietary Official leaderboard | 1683+11/-11 |
| 9 | Gemini 4 Argon (High) (Google) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 9 (rank spread 7-13); 1,775 votes; license: Proprietary Official leaderboard | 1678+15/-15 (Preliminary) |
| 10 | Qwen3.8-Max (0902) (Alibaba) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 10 (rank spread 8-11); 13,812 votes; license: Proprietary Official leaderboard | 1674+7/-7 (Preliminary) |
| 11 | Qwen3.8-Max (Alibaba) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 11 (rank spread 7-15); 3,440 votes; license: Proprietary Official leaderboard | 1672+12/-12 (Preliminary) |
| 12 | Muse Spark 1.3 (Max) (Meta) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 12 (rank spread 10-15); 9,068 votes; license: Proprietary Official leaderboard | 1658+8/-8 |
| 13 | Claude Opus 5 (High) (Anthropic) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 13 (rank spread 11-15); 24,211 votes; license: Proprietary Official leaderboard | 1657+5/-5 |
| 14 | Kimi K3 (Max) (Moonshot) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 14 (rank spread 11-15); 19,220 votes; license: Kimi K3 license Official leaderboard | 1655+6/-6 |
| 15 | Muse Wonder (Max) (Meta) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 15 (rank spread 10-19); 1,942 votes; license: Proprietary Official leaderboard | 1650+15/-15 |
| 16 | Grok 4.7 (xHigh) (SpaceXAI) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 16 (rank spread 15-23); 5,021 votes; license: Proprietary Official leaderboard | 1639+10/-10 |
| 17 | MiMo-V2.6-Flash (Xiaomi) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 17 (rank spread 15-25); 5,016 votes; license: MIT Official leaderboard | 1634+10/-10 |
| 18 | Hy4 preview (Tencent) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 18 (rank spread 15-26); 4,802 votes; license: Apache 2.0 Official leaderboard | 1631+10/-10 (Preliminary) |
| 19 | Qwen3.8-Flash-Next (Alibaba) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 19 (rank spread 15-25); 8,083 votes; license: qwen-community-1.0 Official leaderboard | 1630+8/-8 (Preliminary) |
| 20 | MiMo-V2.6-Pro (Xiaomi) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 20 (rank spread 16-26); 4,763 votes; license: MIT Official leaderboard | 1625+10/-10 |
| 21 | Claude Fable 5 (High) (Anthropic) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 21 (rank spread 16-26); 15,612 votes; license: Proprietary Official leaderboard | 1625+6/-6 |
| 22 | Muse Spark 1.3 (xHigh) (Meta) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 22 (rank spread 16-26); 8,436 votes; license: Proprietary Official leaderboard | 1622+8/-8 |
| 23 | GLM-5.3 (Max) (Z.ai) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 23 (rank spread 17-26); 9,542 votes; license: MIT Official leaderboard | 1621+8/-8 |
| 24 | DeepSeek-V4.1-Flash (Max) (DeepSeek) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 24 (rank spread 16-28); 3,540 votes; license: MIT Official leaderboard | 1618+11/-11 |
| 25 | Grok 4.6 (High) (SpaceXAI) Code Arena WebDev Overall, default leaderboard view (891,922 votes, 145 models); rank 25 (rank spread 17-28); 10,179 votes; license: Proprietary Official leaderboard | 1617+7/-7 |
About this board
- The two boards have separate score scales, so their scores cannot be compared with each other.
FAQ
What does LMArena measure?
LMArena ranks AI models on two boards: Text, covering math, coding, creative writing and other open-ended tasks, and WebDev, covering front-end web development including multi-step agentic coding with tool use. This page covers 2 separate boards: LMArena Text (Text Arena Overall) and LMArena WebDev (Code Arena WebDev Overall). Each is scored on its own terms, so compare rows only within the same board.
Which model leads LMArena?
As of October 17, 2026, Gemini 4 Argon (High) (Google) ranks first on the LMArena Text (Text Arena Overall) board with 1524±9 (Preliminary) and Claude Opus 5.5 (Max) (Anthropic) ranks first on the LMArena WebDev (Code Arena WebDev Overall) board with 1807+14/-14. That is the board's own order, not a Developers Digest ranking.
Where do these LMArena scores come from?
All 50 entries come from the official leaderboard at arena.ai. Every entry links to its source, which was read on October 17, 2026, and lists the setting it was run under.
Related reading
- Grok Imagine Image 2.0 Ships: xAI's Typography-Aware Image Model Is Already on Vercel's AI GatewayGrok Imagine Image 2.0 is xAI's typography-aware image model, ranked second on text-to-image and editing leaderboards and callable via Vercel's AI Gateway.
- Web Dev Arena: How to Test AI Coding Models on Real Frontend WorkWeb Dev Arena compares AI coding models on real frontend work: same prompt, different models, live output, plus frontend evals worth running beyond benchmarks.
- xAI Grok 3 Launch: The Smartest AI on Earth?xAI's Grok 3 launch explained: the 200,000-GPU cluster, benchmarks against GPT-4o and Sonnet 3.5, the reasoning modes, and what it all means for developers.
Sources
Last verified . Each row links to the page its score was copied from.
Newsletter
Get new benchmark results as they land
One email per week. New model releases, the leaderboard moves that matter and what they mean for what you build.