LMArena Text Arena leaderboard (Nov 2026): AI model scores
As of 11 Nov 2026, the highest reported score on LMArena Text Arena is Gemini 4 Carbon (High) (Google) at 1526 (Arena score), per LMArena.
Source: LMArena leaderboard ↗ · Leaderboard dated 11 Nov 2026 · Last checked 17 Nov 2026
LMArena Text Arena
Rankings from human votes on text tasks across maths, coding, creative writing and other open-ended domains.
LMArena publishes per-token prices for some models, not a cost per test, so token prices would not be a fair cost comparison. Showing the ranking instead.
Highest reported score: Gemini 4 Carbon (High) (Google) at 1526.
Arena score, highest first. Position is the source’s published rank.
Bars start at 1470, not zero, so small gaps look larger.
- 1Gemini 4 Carbon (High) Google1526
- 2Gemini 4 Argon (High) Google1522
- 3Muse Wonder (Max) Meta1511
- 4Claude Opus 5.5 (High) Anthropic1507
- 5Claude Fable 5 (High) Anthropic1504
- 6Claude Opus 4.6 (High) Anthropic1503
- 7Claude Fable 5.1 (Max) Anthropic1502
- 8Claude Opus 4.7 (High) Anthropic1501
- 9Gemini 3.8 Flash (High) Google1497
- 10Muse Spark 1.3 (Max) Meta1497
- 11Claude Opus 4.6 Anthropic1496
- 12Claude Opus 4.7 Anthropic1494
- 13Muse Spark 1.2 (xHigh) Meta1493
- 14Muse Spark 1.1 Meta1491
- 15Gemini 4 Hydrogen (High) Google1490
- 16Claude Opus 5 (High) Anthropic1490
- 17Gemini 3.7 Flash (High) Google1489
- 18Muse Spark Meta1489
- 19Claude Opus 5 (Max) Anthropic1489
- 20Kimi K3 (Max) Moonshot1487
- 21Gemini 3.1 Pro Preview Google1487
- 22Gemini 3 Pro Google1486
- 23GPT-5.6 Sol (xHigh) OpenAI1485
- 24Kimi K3.1 Pro (Max) Moonshot1485
- 25Qwen3.8-Max Alibaba1484
- 28GLM-5.3 (Max) Z.ai1482
- 29GPT-5.5 (High) OpenAI1481
- 30GPT-6.1 Sol (Max) OpenAI1481
- 38DeepSeek-V4.1-Flash-20260910 (Max) DeepSeek1475
- 40Grok 4.20 Beta 1 xAI1475
| 1 | Gemini 4 Carbon (High) | 1526 | $10 | |
|---|---|---|---|---|
| 2 | Gemini 4 Argon (High) | 1522 | $10 | |
| 3 | Muse Wonder (Max) | Meta | 1511 | $20 |
| 4 | Claude Opus 5.5 (High) | Anthropic | 1507 | $20 |
| 5 | Claude Fable 5 (High) | Anthropic | 1504 | $50 |
| 6 | Claude Opus 4.6 (High) | Anthropic | 1503 | $25 |
| 7 | Claude Fable 5.1 (Max) | Anthropic | 1502 | $50 |
| 8 | Claude Opus 4.7 (High) | Anthropic | 1501 | $25 |
| 9 | Gemini 3.8 Flash (High) | 1497 | $3.75 | |
| 10 | Muse Spark 1.3 (Max) | Meta | 1497 | $4.25 |
| 11 | Claude Opus 4.6 | Anthropic | 1496 | $25 |
| 12 | Claude Opus 4.7 | Anthropic | 1494 | $25 |
| 13 | Muse Spark 1.2 (xHigh) | Meta | 1493 | $4.25 |
| 14 | Muse Spark 1.1 | Meta | 1491 | $4.25 |
| 15 | Gemini 4 Hydrogen (High) | 1490 | $7.50 | |
| 16 | Claude Opus 5 (High) | Anthropic | 1490 | $25 |
| 17 | Gemini 3.7 Flash (High) | 1489 | $3.75 | |
| 18 | Muse Spark | Meta | 1489 | – |
| 19 | Claude Opus 5 (Max) | Anthropic | 1489 | $25 |
| 20 | Kimi K3 (Max) | Moonshot | 1487 | – |
| 21 | Gemini 3.1 Pro Preview | 1487 | $6.00 | |
| 22 | Gemini 3 Pro | 1486 | $12 | |
| 23 | GPT-5.6 Sol (xHigh) | OpenAI | 1485 | $20 |
| 24 | Kimi K3.1 Pro (Max) | Moonshot | 1485 | $10 |
| 25 | Qwen3.8-Max | Alibaba | 1484 | $6.00 |
| 28 | GLM-5.3 (Max) | Z.ai | 1482 | $4.40 |
| 29 | GPT-5.5 (High) | OpenAI | 1481 | $30 |
| 30 | GPT-6.1 Sol (Max) | OpenAI | 1481 | $10 |
| 38 | DeepSeek-V4.1-Flash-20260910 (Max) | DeepSeek | 1475 | $1.20 |
| 40 | Grok 4.20 Beta 1 | xAI | 1475 | – |
Source: LMArena · leaderboard dated 11 Nov 2026 · checked 17 Nov 2026
Prompts to try with these models
All prompts →Free to use on your site, blog or newsletter. Please keep the “via chatjudge.ai” credit and the source link under the chart. The chart updates itself when we refresh the scores.
Citation
LMArena Text Arena, arena score: Gemini 4 Carbon (High) (Google) had the top reported score at 1526. Original source: LMArena, https://lmarena.ai/leaderboard/text, leaderboard dated 11 Nov 2026. Via ChatJudge: https://chatjudge.ai/benchmarks/chat/lmarena-text, checked 17 Nov 2026.
Other published views: Score vs price
- Anthropic
- OpenAI
- xAI
- Meta
- DeepSeek
- Alibaba
- Z.ai
- Moonshot
Methodology and caveats
What it measures
Rankings from human votes on text tasks across maths, coding, creative writing and other open-ended domains.
People chat with two anonymous models side by side and vote for the better answer; votes become an Elo-style score.
What the score means
Arena score: rating from head-to-head human votes
- Arena score
- Elo-style rating from 9.0M+ head-to-head human votes (± is the confidence interval).
- Output price ($ per 1M tokens)
- List price for output tokens, as shown on the leaderboard.
Run by LMArena. Leaderboard dated 11 Nov 2026. Last checked 17 Nov 2026. Showing top 30 of 420 models listed.
Top 25 overall, plus the best-ranked entry for OpenAI, xAI, DeepSeek and Z.ai (ranks 26–39). Models without a listed price are left off the price view.
Source: LMArena leaderboard ↗
Results from published AI leaderboards. Compare models within each benchmark; scores from different tests aren’t directly comparable. How we collect these results
Top model here: Gemini 4 Carbon (High) (Google). Try prompts for Gemini →
Dated snapshots
Permanent copies of this leaderboard, one per board date, each with the test variant, source date and a CSV.
Cited in public research
4 public reports from Forecasting Agent (Agentic Learning AI Lab, New York University) used this page to check LMArena Text Arena results. The scores come from LMArena; ChatJudge copies them with the leaderboard date. LMArena leaderboard ↗
Forecasting Agent · report 7128 ·
Question: Will no company have an AI model hit 1550 on Chatbot Arena in 2026?
Used ChatJudge to check the top Text Arena scores on 27–28 Oct 2026 (Gemini 4 Carbon 1528, Gemini 4 Argon 1520), to track how close any model was to 1550. Read the passage in the public report ↗ · See the board dated 27 Oct 2026
Forecasting Agent · report 7020 ·
Question: Will Google have the best AI model at the end of December 2026?
Used ChatJudge to check Gemini 4 Carbon’s 1528 on the 27–28 Oct 2026 board, to show its lead was holding. Read the passage in the public report ↗ · See the board dated 27 Oct 2026
Forecasting Agent · report 6444 ·
Question: Will Google have the best AI model at the end of December 2026?
Used ChatJudge to check that Google held ranks 1 and 2 on the 27 Oct 2026 board, with the best non-Google model at #3. Read the passage in the public report ↗ · See the board dated 27 Oct 2026
Forecasting Agent · report 6396 ·
Question: Will Google have the best AI model at the end of November 2026?
Used ChatJudge to check who led Text Arena on the 27 Oct 2026 board: Google’s Gemini 4 Carbon at 1528, 19 points ahead of Claude Opus 5.5 (1509). Read the passage in the public report ↗ · See the board dated 27 Oct 2026
Questions people ask
Which AI has the highest reported score on LMArena Text Arena?
As of 11 Nov 2026, the highest reported score on LMArena Text Arena is Gemini 4 Carbon (High) (Google) at 1526 (Arena score), per LMArena.
What does LMArena Text Arena measure?
Rankings from human votes on text tasks across maths, coding, creative writing and other open-ended domains. It is run by LMArena.
What does the LMArena Text Arena score mean?
Arena score: rating from head-to-head human votes. Elo-style rating from 9.0M+ head-to-head human votes (± is the confidence interval).
How recent are these LMArena Text Arena results?
LMArena Text Arena: leaderboard dated 11 Nov 2026; we last checked it on 17 Nov 2026.
Can I compare LMArena Text Arena scores with other benchmarks?
Not directly. Each leaderboard uses its own tasks, settings and scoring, so compare models within one benchmark. ChatJudge does not run the tests; we copy the published scores with a link to the source.
