LMArena Hard Prompts leaderboard
As of 15 Nov 2026, the top score on LMArena Hard Prompts is 1547 by Gemini 4 Carbon (Google), ahead of Gemini 4 Argon at 1545, across 384 model configurations with a published result (Arena (LMArena), result dated 11 Nov 2026). The same pairwise votes as the text arena, restricted to prompts that LMArena's classifier judges hard: specific, complex, multi-step requests rather than casual chat. LMArena Hard Prompts counts toward the BenchLeader reasoning score.
Key facts (as of 15 Nov 2026)
- Gemini 4 Carbon holds the top Arena Hard Prompts score at 1547 (Arena (LMArena), 11 Nov 2026).
- Gemini 4 Argon is second on Arena Hard Prompts at 1545 (Arena (LMArena), 11 Nov 2026).
- Arena Hard Prompts: Text arena rating on prompts judged hard (Arena (LMArena)).
- 384 model configurations have a published Arena Hard Prompts result on BenchLeader (BenchLeader, 15 Nov 2026).
- Arena Hard Prompts feeds the reasoning category of the BenchLeader Index (BenchLeader methodology).
Text arena rating on prompts judged hard.
- Published by
- Arena (LMArena)
- Category
- Reasoning
- Index weight
- 0.5
- Models
- 384
- Data as of
- 15 Nov 2026
This week on Arena Hard Prompts
8 Nov 2026 to 15 Nov 2026, one row per model (its best setting on this board).
- Up: Muse Spark 1.3 #12 → #10
- Down: Kimi K3 #10 → #11, Claude Opus 5 #11 → #12, Muse Spark 1.2 #13 → #14, Claude Opus 4.8 #14 → #15, GPT-5.6 Sol #15 → #16, Muse Spark 1.1 #16 → #17, Gemini 3.7 Flash #17 → #18
- New to the top 20: Gemini 4 Hydrogen at #13 (1514)
CC BY 4.0 — lmarena-ai/leaderboard-dataset on Hugging Face.
What the test looks like
The same pairwise votes as the text arena, restricted to prompts that LMArena's classifier judges hard: specific, complex, multi-step requests rather than casual chat.
How it is scored
Bradley-Terry rating on the hard-prompt subset, with style control.
What to keep in mind
Fewer votes per model than the overall board, so confidence intervals are wider. Hardness is judged by a classifier, not by outcome.
- 1Gemini 4 Carbon (high)1547
- 2Gemini 4 Argon (high)1545
- 3Muse Wonder (max)1531
- 4Claude Opus 5.5 (high)1530
- 5Claude Opus 4.6 (high)1530
- 6Claude Fable 5 (high)1530
- 7Claude Opus 4.61526
- 8Claude Opus 4.7 (high)1525
- 9Claude Fable 5.1 (max)1522
- 10Claude Opus 4.71519
- 11Gemini 3.8 Flash (high)1518
- 12Muse Spark 1.3 (max)1517
- 13Kimi K3 (max)1516
- 14Claude Opus 5 (high)1515
- 15Gemini 4 Hydrogen (high)1514
| # | |||||
|---|---|---|---|---|---|
| 1 | 1547 | 68.9 | 2026-11-11 | $4.00 | |
| 2 | 1545 | 65.3 | 2026-11-11 | $4.00 | |
| 3 | 1531 | 60.8 | 2026-11-11 | $8.00 | |
| 4 | 1530 | 66.0 | 2026-11-11 | $8.00 | |
| 5 | 1530 | 62.5 | 2026-11-11 | $10.00 | |
| 6 | 1530 | 63.9 | 2026-11-11 | $20.00 | |
| 7 | 1526 | 62.1 | 2026-11-11 | $10.00 | |
| 8 | 1525 | 60.0 | 2026-11-11 | $10.00 | |
| 9 | 1522 | 66.7 | 2026-11-11 | $20.00 | |
| 10 | 1519 | 63.1 | 2026-11-11 | $10.00 | |
| 11 | 1518 | 59.4 | 2026-11-11 | $1.50 | |
| 12 | 1517 | 60.6 | 2026-11-11 | $2.00 | |
| 13 | 1516 | 60.2 | 2026-11-11 | $6.00 | |
| 14 | 1515 | 64.4 | 2026-11-11 | $10.00 | |
| 15 | New | 1514 | 63.4 | 2026-11-11 | $3.00 |
| 16 | 1514 | 61.5 | 2026-11-11 | $2.00 | |
| 17 | 1513 | 65.0 | 2026-11-11 | $10.00 | |
| 18 | 1513 | 60.4 | 2026-11-11 | $10.00 | |
| 19 | 1511 | 63.9 | 2026-11-11 | $8.00 | |
| 20 | 1511 | 62.2 | 2026-11-11 | $2.00 | |
| 21 | 1510 | 62.1 | 2026-11-11 | $1.50 | |
| 22 | 1508 | 59.9 | 2026-11-11 | $4.50 | |
| 23 | 1507 | 58.8 | 2026-11-11 | $2.15 | |
| 24 | 1507 | 58.7 | 2026-11-11 | $3.00 | |
| 25 | 1506 | 64.2 | 2026-11-11 | – | |
| 26 | 1505 | 58.7 | 2026-11-11 | $10.00 | |
| 27 | 1505 | 66.6 | 2026-11-11 | $4.00 | |
| 28 | 1504 | 57.5 | 2026-11-11 | $0.548 | |
| 29 | 1504 | 59.9 | 2026-11-11 | $4.00 | |
| 30 | 1503 | 57.6 | 2026-11-11 | $6.00 | |
| 31 | 1502 | 60.0 | 2026-11-11 | – | |
| 32 | 1501 | 61.0 | 2026-11-11 | $1.50 | |
| 33 | 1501 | 63.5 | 2026-11-11 | $11.25 | |
| 34 | 1501 | 59.7 | 2026-11-11 | $10.00 | |
| 35 | 1500 | 57.8 | 2026-11-11 | $0.238 | |
| 36 | 1500 | 67.3 | 2026-11-11 | $20.00 | |
| 37 | 1499 | 62.5 | 2026-11-11 | $4.00 | |
| 38 | 1499 | 61.0 | 2026-11-11 | $11.25 | |
| 39 | 1499 | 58.5 | 2026-11-11 | $0.525 | |
| 40 | 1499 | 57.1 | 2026-11-11 | $10.00 | |
| 41 | 1497 | 56.7 | 2026-11-11 | $4.81 | |
| 42 | 1496 | 58.6 | 2026-11-11 | $5.63 | |
| 43 | 1496 | 62.5 | 2026-11-11 | $3.38 | |
| 44 | 1496 | 58.9 | 2026-11-11 | $2.15 | |
| 45 | 1495 | 61.8 | 2026-11-11 | $3.38 | |
| 46 | 1495 | 60.0 | 2026-11-11 | $3.75 | |
| 47 | 1494 | 55.5 | 2026-11-11 | $1.50 | |
| 48 | 1493 | 56.5 | 2026-11-11 | $1.13 | |
| 49 | 1492 | – | 2026-11-11 | – | |
| 50 | 1491 | 56.9 | 2026-11-11 | $11.25 | |
| 51 | 1491 | 60.1 | 2026-11-11 | $4.50 | |
| 52 | 1490 | 59.0 | 2026-11-11 | $4.00 | |
| 53 | 1489 | 58.3 | 2026-11-11 | $1.56 | |
| 54 | 1489 | 55.4 | 2026-11-11 | $2.15 | |
| 55 | 1488 | 59.3 | 2026-11-11 | $3.00 | |
| 56 | 1487 | – | 2026-11-11 | – | |
| 57 | 1487 | – | 2026-11-11 | – | |
| 58 | 1487 | 57.9 | 2026-11-11 | $5.63 | |
| 59 | 1487 | 57.9 | 2026-11-11 | $6.00 | |
| 60 | 1485 | 55.1 | 2026-11-11 | $1.71 |
AI reasoning and math benchmark leaderboard · All LLM benchmarks
Related benchmarks
Top head-to-heads on Arena Hard Prompts
Cite as: BenchLeader, “Arena Hard Prompts leaderboard”, data as of 15 Nov 2026, https://www.benchleader.com/benchmarks/lmarena_text_hard.
Arena Hard Prompts: questions
- What does Arena Hard Prompts measure?
- The same pairwise votes as the text arena, restricted to prompts that LMArena's classifier judges hard: specific, complex, multi-step requests rather than casual chat. Scores are reported in Elo-style ratings from pairwise comparisons; higher is better.
- What is the top score on LMArena Hard Prompts?
- As of 15 Nov 2026, the highest score on Arena Hard Prompts is 1547 by Gemini 4 Carbon (result dated 11 Nov 2026), ahead of Gemini 4 Argon at 1545. Scores come from Arena (LMArena) and are re-read daily.
- Which AI model leads Arena Hard Prompts?
- Gemini 4 Carbon leads Arena Hard Prompts with 1547 as of 15 Nov 2026 (result dated 11 Nov 2026), ahead of Gemini 4 Argon at 1545.
- How many models have Arena Hard Prompts results?
- 384 model configurations have a Arena Hard Prompts result on BenchLeader, all taken from Arena (LMArena).
- Who runs Arena Hard Prompts and how often is it updated?
- Arena Hard Prompts is published by Arena (LMArena). BenchLeader re-reads the published results once a day and records the date each result was published.
- Does Arena Hard Prompts count toward the BenchLeader Index?
- Yes. Arena Hard Prompts contributes to the reasoning category of the BenchLeader Index, normalised so that 50 is the average of the evaluated models.