0.02
Confirmed task success · rank 16 of 46
- Unit
- IPS effect estimate, higher is better
- Range
- 0.01 to 0.04
- Sample
- 35046 observations
- Configuration
- Qwen3.8 27B
- Measured
- 28 Sep 2026
- Not shown
- Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.
-0.03
Praise over complaint · rank 21 of 46
- Unit
- IPS effect estimate, higher is better
- Range
- -0.05 to -0.01
- Sample
- 14131 observations
- Configuration
- Qwen3.8 27B
- Measured
- 28 Sep 2026
- Not shown
- Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.
-0.01
Steerability · rank 21 of 46
- Unit
- IPS effect estimate, higher is better
- Range
- -0.02 to 0.00
- Sample
- 40165 observations
- Configuration
- Qwen3.8 27B
- Measured
- 28 Sep 2026
- Not shown
- Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.
-0.00
Tool grounding · rank 31 of 46
- Unit
- IPS effect estimate, higher is better
- Range
- -0.00 to -0.00
- Sample
- 5969375 observations
- Configuration
- Qwen3.8 27B
- Measured
- 28 Sep 2026
- Not shown
- Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.
1,439
Overall · rank 69 of 177
- Unit
- Arena rating, higher is better
- Range
- 1,434 to 1,443
- Sample
- 16057 votes
- Configuration
- Qwen3.8 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.
1,434
Business, management and finance · rank 65 of 402
- Unit
- Arena rating, higher is better
- Range
- 1,423 to 1,445
- Sample
- 3087 votes
- Configuration
- Qwen3.8 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,364
Creative writing · rank 128 of 407
- Unit
- Arena rating, higher is better
- Range
- 1,353 to 1,375
- Sample
- 3220 votes
- Configuration
- Qwen3.8 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,477
Expert prompts · rank 33 of 359
- Unit
- Arena rating, higher is better
- Range
- 1,463 to 1,491
- Sample
- 1808 votes
- Configuration
- Qwen3.8 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,423
Instruction following · rank 77 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,415 to 1,432
- Sample
- 5819 votes
- Configuration
- Qwen3.8 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,438
Overall · rank 79 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,432 to 1,443
- Sample
- 16383 votes
- Configuration
- Qwen3.8 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,396
Writing, literature and language · rank 94 of 408
- Unit
- Arena rating, higher is better
- Range
- 1,386 to 1,406
- Sample
- 4262 votes
- Configuration
- Qwen3.8 27B
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
