1,380
Overall · rank 149 of 177
- Unit
- Arena rating, higher is better
- Range
- 1,369 to 1,391
- Sample
- 2377 votes
- Configuration
- Gemma 3 27B
- 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.
1,377
Business, management and finance · rank 161 of 402
- Unit
- Arena rating, higher is better
- Range
- 1,369 to 1,385
- Sample
- 6472 votes
- Configuration
- Gemma 3 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,348
Creative writing · rank 153 of 407
- Unit
- Arena rating, higher is better
- Range
- 1,341 to 1,356
- Sample
- 6271 votes
- Configuration
- Gemma 3 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,337
Expert prompts · rank 201 of 359
- Unit
- Arena rating, higher is better
- Range
- 1,325 to 1,350
- Sample
- 2237 votes
- Configuration
- Gemma 3 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,343
Instruction following · rank 194 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,337 to 1,349
- Sample
- 12440 votes
- Configuration
- Gemma 3 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,365
Overall · rank 188 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,362 to 1,369
- Sample
- 46433 votes
- Configuration
- Gemma 3 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,354
Writing, literature and language · rank 164 of 408
- Unit
- Arena rating, higher is better
- Range
- 1,348 to 1,361
- Sample
- 10536 votes
- Configuration
- Gemma 3 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
73.7%
Irrelevance detection · rank 80 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not general refusal or safety behaviour.
13.6%
Memory · rank 62 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not long-term personal memory in a product: sessions are BFCL's scripted ones.
10.8%
Multi-turn tasks · rank 66 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not open-ended agent work: the tools and tasks are BFCL's simulated APIs.
29.5%
Overall accuracy · rank 69 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not a neutral average: the weighting is BFCL's. Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
81.2%
Relevance detection · rank 41 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not whether the call itself was right; only that one was attempted.
87.2%
Single-turn calls (curated) · rank 28 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
74.5%
Single-turn calls (user-contributed) · rank 44 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
0.0%
Web search · rank 92 of 109
- Unit
- % correct, higher is better
- Configuration
- gemma-3-27b-it
- Measured
- 16 Dec 2025
- Not shown
- Not general research quality: questions have short, checkable answers.
8.0%
Requested-length error · rank 55 of 370
- Unit
- % off the requested word count, lower is better
- Configuration
- Gemma 3 27B
- Measured
- 6 Sep 2025
- Not shown
- Not other format limits such as character counts or bullet counts.
0.29
Style adherence · rank 343 of 370
- Unit
- score from 0 to 1, higher is better
- Configuration
- Gemma 3 27B
- Measured
- 6 Sep 2025
- Not shown
- Not brand voice on your own examples: UGI's prompts are private and lean towards creative writing.
45.0
Writing score · rank 177 of 370
- Unit
- score out of 100, higher is better
- Configuration
- Gemma 3 27B
- Measured
- 6 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.
98.8%
Answer rate · rank 72 of 108
- Unit
- % of documents, higher is better
- Configuration
- Gemma 3 27B
- 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.
7.4%
Hallucination rate · rank 35 of 108
- Unit
- % of summaries, lower is better
- Configuration
- Gemma 3 27B
- 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.
