1,385
Overall · rank 149 of 177
- Unit
- Arena rating, higher is better
- Range
- 1,378 to 1,391
- Sample
- 7347 votes
- Configuration
- gpt-oss-120b
- 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.
1,351
Business, management and finance · rank 184 of 402
- Unit
- Arena rating, higher is better
- Range
- 1,343 to 1,360
- Sample
- 5339 votes
- Configuration
- gpt-oss-120b
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,277
Creative writing · rank 229 of 407
- Unit
- Arena rating, higher is better
- Range
- 1,267 to 1,287
- Sample
- 3864 votes
- Configuration
- gpt-oss-120b
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,357
Expert prompts · rank 180 of 359
- Unit
- Arena rating, higher is better
- Range
- 1,341 to 1,373
- Sample
- 1370 votes
- Configuration
- gpt-oss-120b
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,323
Instruction following · rank 209 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,316 to 1,330
- Sample
- 7793 votes
- Configuration
- gpt-oss-120b
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,352
Overall · rank 200 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,347 to 1,356
- Sample
- 30021 votes
- Configuration
- gpt-oss-120b
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,309
Writing, literature and language · rank 215 of 408
- Unit
- Arena rating, higher is better
- Range
- 1,301 to 1,317
- Sample
- 6581 votes
- Configuration
- gpt-oss-120b
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
107.0%
Requested-length error · rank 353 of 370
- Unit
- % off the requested word count, lower is better
- Configuration
- gpt-oss-120b
- Measured
- 1 Oct 2025
- Not shown
- Not other format limits such as character counts or bullet counts.
0.29
Style adherence · rank 351 of 370
- Unit
- score from 0 to 1, higher is better
- Configuration
- gpt-oss-120b
- Measured
- 1 Oct 2025
- Not shown
- Not brand voice on your own examples: UGI's prompts are private and lean towards creative writing.
38.5
Writing score · rank 226 of 370
- Unit
- score out of 100, higher is better
- Configuration
- gpt-oss-120b
- Measured
- 1 Oct 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.
99.9%
Answer rate · rank 19 of 108
- Unit
- % of documents, higher is better
- Configuration
- gpt-oss-120b
- 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.
14.2%
Hallucination rate · rank 90 of 108
- Unit
- % of summaries, lower is better
- Configuration
- gpt-oss-120b
- 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.
