1,431
Overall · rank 34 of 44
- Unit
- Arena rating, higher is better
- Range
- 1,424 to 1,437
- Sample
- 25110 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 13 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,142
Overall · rank 27 of 34
- Unit
- Arena rating, higher is better
- Range
- 1,137 to 1,147
- Sample
- 83404 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 24 Aug 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,462
Overall · rank 28 of 177
- Unit
- Arena rating, higher is better
- Range
- 1,459 to 1,465
- Sample
- 69736 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 25 Sep 2026
- Not shown
- Not an error rate: claims that cannot be checked on the web are skipped, and preference still carries most of the weight.
1,437
Business, management and finance · rank 69 of 402
- Unit
- Arena rating, higher is better
- Range
- 1,433 to 1,442
- Sample
- 22770 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,443
Creative writing · rank 26 of 407
- Unit
- Arena rating, higher is better
- Range
- 1,438 to 1,448
- Sample
- 17590 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,458
Expert prompts · rank 76 of 359
- Unit
- Arena rating, higher is better
- Range
- 1,451 to 1,465
- Sample
- 8260 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,438
Instruction following · rank 61 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,434 to 1,442
- Sample
- 35043 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,446
Overall · rank 72 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,443 to 1,448
- Sample
- 124887 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,445
Writing, literature and language · rank 39 of 408
- Unit
- Arena rating, higher is better
- Range
- 1,440 to 1,449
- Sample
- 28118 votes
- Configuration
- Gemini 2.5 Pro
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1.0%
Consistency (pass^4) · rank 21 of 21
- Unit
- % of tasks, higher is better
- Configuration
- gemini-2.5-pro
- Measured
- 5 May 2026
- Not shown
- Not your policies or customers: the tasks and policies are tau2's own, and the customer is simulated by gpt-5.2.
13.7%
Task success (pass^1) · rank 20 of 21
- Unit
- % of tasks, higher is better
- Configuration
- gemini-2.5-pro
- Measured
- 5 May 2026
- Not shown
- Not your policies or customers: the tasks and policies are tau2's own, and the customer is simulated by gpt-5.2.
19.0%
Requested-length error · rank 163 of 370
- Unit
- % off the requested word count, lower is better
- Configuration
- Gemini 2.5 Pro
- Measured
- 9 Sep 2025
- Not shown
- Not other format limits such as character counts or bullet counts.
0.40
Style adherence · rank 29 of 370
- Unit
- score from 0 to 1, higher is better
- Configuration
- Gemini 2.5 Pro
- Measured
- 9 Sep 2025
- Not shown
- Not brand voice on your own examples: UGI's prompts are private and lean towards creative writing.
65.2
Writing score · rank 63 of 370
- Unit
- score out of 100, higher is better
- Configuration
- Gemini 2.5 Pro
- Measured
- 9 Sep 2025
- Not shown
- Not business copy quality: UGI's prompts are private and lean towards creative writing, and models that often refuse get no score.
99.1%
Answer rate · rank 64 of 108
- Unit
- % of documents, higher is better
- Configuration
- Gemini 2.5 Pro
- Measured
- 22 Sep 2026
- Not shown
- Not a quality score: a low rate usually means content filters were triggered, and hallucination rates are measured on answered documents only.
7.0%
Hallucination rate · rank 32 of 108
- Unit
- % of summaries, lower is better
- Configuration
- Gemini 2.5 Pro
- Measured
- 22 Sep 2026
- Not shown
- Not errors in open questions or other tasks: only summarisation, judged by Vectara's own model (HHEM-2.3), not by people, on news-style documents rather than your data.
