# Corp Speak Arena

> Kaggle's Corp Speak Arena has models rewrite 100 everyday sentences in corporate language, and people vote for the better rewrite.

Canonical page: https://benchlm.ai/benchmarks/corp-speak-arena

- Category: [external](/external)
- Last updated: Latest evaluation November 10, 2026

Evaluation results published by [Kaggle](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena). undefined

A response qualifies if it is a plain-text answer of 5 to 100 words. It is a test of tone and style transfer in business writing. Kaggle shows two anonymous outputs for the same prompt and a person picks the better one. The Arena column is the relative rating Kaggle fits to those votes, published with an interval; it is not a percentage.

The rating moves as votes arrive, and Kaggle publishes no confidence level for its interval. Rows with few votes have wide intervals, so neighbouring models are often not separable. Voters are self-selected Kaggle visitors. The board declares no license; we keep published aggregates with attribution and no task or output contents.

A rating gap smaller than the two rows' intervals is not a reliable difference.

- [Kaggle arena leaderboard](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena)
- [Published leaderboard data](https://www.kaggle.com/api/v1/benchmarks/kaggle/corp-speak-arena/leaderboard)

## About Corp Speak Arena

- Year: 2026
- Tasks: 100 plain sentences in 10 categories, rated by human votes
- Format: Relative rating fitted to human pairwise votes
- Difficulty: Tone and style rewriting

A response qualifies if it is a plain-text answer of 5 to 100 words. It is a test of tone and style transfer in business writing. Kaggle shows two anonymous outputs for the same prompt and a person picks the better one. The Arena column is the relative rating Kaggle fits to those votes, published with an interval; it is not a percentage. These results are display-only and do not enter overall or category rankings.

Corp Speak Arena is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (24 source rows)

| Rank | Model | Configuration | Creator | Score |
|------|-------|---------------|---------|-------|
| 1 | [GLM-5.3](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 15 votes; ±186; Evaluated 2026-11-10 | — | 1731 |
| 2 | [Claude Opus 5](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 9 votes; ±220; Evaluated 2026-11-10 | — | 1696 |
| 3 | [Claude Opus 4.7](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 8 votes; ±212; Evaluated 2026-11-10 | — | 1644 |
| 4 | [Claude Opus 5.5](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 21 votes; ±157; Evaluated 2026-11-10 | — | 1616 |
| 5 | [Gemini 3.1 Pro Preview](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 14 votes; ±182; Evaluated 2026-11-10 | — | 1578 |
| 6 | [Grok 4.7](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Xhigh Reasoning; 18 votes; ±166; Evaluated 2026-11-10 | — | 1565 |
| 7 | [Claude Sonnet 5.5](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 22 votes; ±158; Evaluated 2026-11-10 | — | 1562 |
| 8 | [GPT-6.1 Sol](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 26 votes; ±145; Evaluated 2026-11-10 | — | 1548 |
| 9 | [Gemini 3.8 Flash](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 7 votes; ±225; Evaluated 2026-11-10 | — | 1542 |
| 10 | [Claude Fable 5.1](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 15 votes; ±171; Evaluated 2026-11-10 | — | 1541 |
| 11 | [GPT-6 Astra](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 19 votes; ±161; Evaluated 2026-11-10 | — | 1540 |
| 12 | [Gemini 4 Argon](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 21 votes; ±156; Evaluated 2026-11-10 | — | 1498 |
| 13 | [GPT-5.6 Sol](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 9 votes; ±202; Evaluated 2026-11-10 | — | 1481 |
| 14 | [Claude Sonnet 5](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 10 votes; ±195; Evaluated 2026-11-10 | — | 1462 |
| 15 | [GPT-5 Pro](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 12 votes; ±185; Evaluated 2026-11-10 | — | 1460 |
| 16 | [GPT-5.2](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Xhigh Reasoning; 8 votes; ±208; Evaluated 2026-11-10 | — | 1457 |
| 17 | [GPT-5.6 Terra](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 18 votes; ±165; Evaluated 2026-11-10 | — | 1451 |
| 18 | [GPT-5.5](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) |  GPT-5.5 High Reasoning; 20 votes; ±163; Evaluated 2026-11-10 | — | 1441 |
| 19 | [Claude Opus 4.7](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 6 votes; ±227; Evaluated 2026-11-10 | — | 1437 |
| 20 | [GPT-5.6 Terra](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | GPT 5.6 Terra Max Reasoning; 10 votes; ±200; Evaluated 2026-11-10 | — | 1434 |
| 21 | [Qwen 3.8 Max](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Xhigh Reasoning; 13 votes; ±185; Evaluated 2026-11-10 | — | 1416 |
| 22 | [GPT-5](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | High Reasoning; 6 votes; ±235; Evaluated 2026-11-10 | — | 1367 |
| 23 | [GPT-5.4](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Xhigh Reasoning; 13 votes; ±196; Evaluated 2026-11-10 | — | 1364 |
| 24 | [GPT-6 Luna](https://www.kaggle.com/benchmarks/kaggle/corp-speak-arena) | Max Reasoning; 15 votes; ±202; Evaluated 2026-11-10 | — | 1218 |

## FAQ

### What does Corp Speak Arena measure?

Kaggle's Corp Speak Arena has models rewrite 100 everyday sentences in corporate language, and people vote for the better rewrite.

### Which source row has the highest reported Corp Speak Arena value?

GLM-5.3 has the highest reported Corp Speak Arena value at 1731. A rating gap smaller than the two rows' intervals is not a reliable difference.

### How many models are evaluated on Corp Speak Arena?

The Latest evaluation November 10, 2026 contains 24 AI models.

### Does Corp Speak Arena affect BenchLM's overall score?

Not directly. Corp Speak Arena is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.