LMArena · Coding
We regularly fetch and display public benchmark data: LMArena user voting rankings, Epoch AI test scores, and OpenRouter usage rankings. Each leaderboard shows both its publication date and the date we retrieved it. These scores are not assigned by this site.
How to read this leaderboard
What it measures: A ranking calculated from votes classified as coding-related questions in LMArena text conversations.
How to read the score: A relative score reflecting which models users preferred for coding answers. Interpret small score differences alongside confidence intervals and vote counts.
Caveats: This evaluates preferences in coding conversations, so distinguish it from agent evaluations that actually modify and test repositories. Also check SWE-bench, Terminal-Bench and published hands-on results.
Insights from this ranking
- The 95% confidence intervals for 1st-place gemini-4-argon-high and 2nd-place claude-fable-5-high overlap. This aggregation alone does not clearly establish their order.
- 14 other models have confidence intervals that overlap with the 1st-place model's. Interpret small ranking differences cautiously alongside vote counts.
- Anthropic has the most models among the top 10, with 5.
- The 1st-place model received 1,230 votes, and this leaderboard includes 408 models.
Coding LMArena
Published 2026-10-02 · Retrieved 2026-10-04
Dots show scores; horizontal lines show 95% confidence intervals. Overlapping intervals suggest similar performance. The chart shows only the highest-scoring reasoning setting, such as high or max, for each model. The table shows rankings for all settings. Chat and vision rankings use 'style and length adjustment (style control)', the same default as the LMArena website.
Top ranking by company
- Googlegemini-4-argon-high#1
- Anthropicclaude-fable-5-high#2
- OOpenAIgpt-6-astra-max#7
- Moonshot AIkimi-k3-max#9
- Xiaomimimo-v2.6-pro#10
- Metamuse-spark-1.3-max#11
- DeepSeekdeepseek-v4.1-flash-max#26
- Alibabaqwen3.7-max-preview#27
- ZZ.aiglm-5.3-flash#29
View table (top 50)
| Rank | Model | Developer | Score | 95% confidence interval | Votes |
|---|---|---|---|---|---|
| 1 | #1gemini-4-argon-high | 1560 | 1543–1577 | 1,230 | |
| 2 | #2claude-fable-5-high | Anthropic | 1552 | 1545–1559 | 10,038 |
| 3 | #3claude-opus-4-6-high | Anthropic | 1551 | 1546–1557 | 20,235 |
| 4 | #4claude-opus-4-7-high | Anthropic | 1551 | 1545–1556 | 18,521 |
| 5 | #5claude-opus-4-6 | Anthropic | 1548 | 1542–1553 | 22,901 |
| 6 | #6claude-opus-4-7 | Anthropic | 1546 | 1541–1552 | 18,728 |
| 7 | #7Ogpt-6-astra-max | OpenAI | 1543 | 1530–1556 | 2,135 |
| 8 | #8Ogpt-6.1-sol-max | OpenAI | 1542 | 1518–1566 | 609 |
| 9 | #9kimi-k3-max | Moonshot AI | 1541 | 1533–1549 | 7,285 |
| 10 | #10mimo-v2.6-pro | Xiaomi | 1540 | 1522–1558 | 1,103 |
| 11 | #11muse-spark-1.3-max | Meta | 1539 | 1528–1549 | 3,307 |
| 12 | #12claude-opus-5.5-high | Anthropic | 1538 | 1521–1556 | 1,060 |
| 13 | #13claude-sonnet-5.5-xhigh | Anthropic | 1536 | 1515–1558 | 779 |
| 14 | #14Ogpt-5.6-sol-xhigh | OpenAI | 1534 | 1527–1541 | 9,943 |
| 15 | #15claude-opus-5-high | Anthropic | 1533 | 1527–1540 | 15,559 |
| 16 | #16claude-opus-4-8-high | Anthropic | 1533 | 1527–1539 | 17,395 |
| 17 | #17muse-spark-1.1 | Meta | 1533 | 1526–1539 | 10,537 |
| 18 | #18muse-spark-1.2 (xHigh) | Meta | 1531 | 1514–1549 | 1,222 |
| 19 | #19claude-fable-5.1-max | Anthropic | 1531 | 1519–1544 | 2,358 |
| 20 | #20claude-opus-4-5-20251101-high-32k | Anthropic | 1530 | 1523–1538 | 7,793 |
| 21 | #21claude-opus-5-max | Anthropic | 1530 | 1522–1538 | 7,382 |
| 22 | #22gemini-3.8-flash-high | 1530 | 1522–1538 | 7,067 | |
| 23 | #23claude-opus-4-8 | Anthropic | 1530 | 1524–1535 | 18,024 |
| 24 | #24muse-spark | Meta | 1530 | 1520–1540 | 3,930 |
| 25 | #25claude-sonnet-4-6 | Anthropic | 1529 | 1523–1534 | 19,686 |
| 26 | #26deepseek-v4.1-flash-max | DeepSeek | 1528 | 1516–1540 | 2,469 |
| 27 | #27qwen3.7-max-preview | Alibaba | 1524 | 1506–1542 | 1,165 |
| 28 | #28qwen3.8-max | Alibaba | 1524 | 1516–1532 | 6,611 |
| 29 | #29Zglm-5.3-flash | Z.ai | 1523 | 1515–1532 | 6,157 |
| 30 | #30claude-opus-4-5-20251101 | Anthropic | 1523 | 1518–1528 | 17,962 |
| 31 | #31gemini-3.1-pro-preview | 1522 | 1517–1526 | 33,581 | |
| 32 | #32mimo-v2.5-pro | Xiaomi | 1522 | 1516–1527 | 19,324 |
| 33 | #33mimo-v2.6-flash | Xiaomi | 1521 | 1506–1537 | 1,508 |
| 34 | #34Zglm-5.3-max | Z.ai | 1521 | 1511–1530 | 4,424 |
| 35 | #35Ogpt-5.4-high | OpenAI | 1521 | 1515–1527 | 17,566 |
| 36 | #36Ogpt-6-sol-max | OpenAI | 1520 | 1507–1534 | 2,071 |
| 37 | #37claude-sonnet-4-5-20250929-high-32k | Anthropic | 1519 | 1514–1524 | 19,905 |
| 38 | #38gemini-3.6-flash-high | 1519 | 1512–1526 | 10,350 | |
| 39 | #39claude-sonnet-5-high | Anthropic | 1519 | 1512–1525 | 12,377 |
| 40 | #40Ogpt-5.5-high | OpenAI | 1519 | 1513–1524 | 19,110 |
| 41 | #41gemini-3.7-flash-high | 1518 | 1510–1526 | 6,041 | |
| 42 | #42gemini-3-pro | 1518 | 1511–1525 | 8,787 | |
| 43 | #43kimi-k2.6 | Moonshot AI | 1516 | 1509–1523 | 11,062 |
| 44 | #44Ogpt-5.6-terra-xhigh | OpenAI | 1515 | 1508–1522 | 10,166 |
| 45 | #45Ogpt-5.2-chat-latest-20260210 | OpenAI | 1515 | 1508–1522 | 9,629 |
| 46 | #46xgrok-4.5 | xAI | 1514 | 1507–1521 | 11,055 |
| 47 | #47dola-seed-2.0-pro | ByteDance | 1514 | 1509–1519 | 21,996 |
| 48 | #48qwen3.5-max-preview | Alibaba | 1514 | 1506–1522 | 6,358 |
| 49 | #49Ogpt-5.5-instant | OpenAI | 1513 | 1506–1521 | 7,909 |
| 50 | #50Zglm-5.1 | Z.ai | 1513 | 1507–1518 | 16,060 |
Data: LMArena leaderboard dataset (CC BY 4.0). Changes: Top 50 entries per leaderboard; scores are rounded. https://huggingface.co/datasets/lmarena-ai/leaderboard-datasetCompany logos are trademarks of their respective owners and are used only for identification (icons: Simple Icons).Data: Epoch AI — Capabilities & benchmarking (CC BY 4.0). https://epoch.ai/benchmarksSource: OpenRouter (openrouter.ai/rankings), as of 2026-10-04. Licensed under CC BY 4.0. https://openrouter.ai/rankings