1,450
Overall · rank 19 of 44
- Unit
- Arena rating, higher is better
- Range
- 1,442 to 1,458
- Sample
- 11291 votes
- Configuration
- Kimi K2.6
- Measured
- 13 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,456
Overall · rank 38 of 177
- Unit
- Arena rating, higher is better
- Range
- 1,452 to 1,460
- Sample
- 39923 votes
- Configuration
- Kimi K2.6
- 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.
1,455
Business, management and finance · rank 37 of 402
- Unit
- Arena rating, higher is better
- Range
- 1,447 to 1,462
- Sample
- 7948 votes
- Configuration
- Kimi K2.6
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,433
Creative writing · rank 37 of 407
- Unit
- Arena rating, higher is better
- Range
- 1,425 to 1,441
- Sample
- 6305 votes
- Configuration
- Kimi K2.6
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,503
Expert prompts · rank 13 of 359
- Unit
- Arena rating, higher is better
- Range
- 1,493 to 1,513
- Sample
- 3841 votes
- Configuration
- Kimi K2.6
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,456
Instruction following · rank 33 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,449 to 1,462
- Sample
- 13226 votes
- Configuration
- Kimi K2.6
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,461
Overall · rank 44 of 409
- Unit
- Arena rating, higher is better
- Range
- 1,456 to 1,465
- Sample
- 39971 votes
- Configuration
- Kimi K2.6
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
1,441
Writing, literature and language · rank 40 of 408
- Unit
- Arena rating, higher is better
- Range
- 1,434 to 1,448
- Sample
- 9562 votes
- Configuration
- Kimi K2.6
- Measured
- 25 Sep 2026
- Not shown
- Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
29.3%
Consistency (pass^5) · rank 3 of 5
- Unit
- % of tasks, higher is better
- Configuration
- kimi-k2.6
- Measured
- 25 May 2026
- Not shown
- Not the model alone: it runs inside STATE-Bench's agent loop against simulated users, and versions of the benchmark are not comparable.
45.1%
Customer support task success (pass@1) · rank 5 of 5
- Unit
- % of tasks, higher is better
- Configuration
- kimi-k2.6
- Measured
- 25 May 2026
- Not shown
- Not the model alone: it runs inside STATE-Bench's agent loop against simulated users, and versions of the benchmark are not comparable.
47.9%
Shopping assistant task success (pass@1) · rank 5 of 5
- Unit
- % of tasks, higher is better
- Configuration
- kimi-k2.6
- Measured
- 25 May 2026
- Not shown
- Not the model alone: it runs inside STATE-Bench's agent loop against simulated users, and versions of the benchmark are not comparable.
48.3%
Task success (pass@1) · rank 3 of 5
- Unit
- % of tasks, higher is better
- Configuration
- kimi-k2.6
- Measured
- 25 May 2026
- Not shown
- Not the model alone: it runs inside STATE-Bench's agent loop against simulated users, and versions of the benchmark are not comparable.
51.9%
Travel task success (pass@1) · rank 3 of 5
- Unit
- % of tasks, higher is better
- Configuration
- kimi-k2.6
- Measured
- 25 May 2026
- Not shown
- Not the model alone: it runs inside STATE-Bench's agent loop against simulated users, and versions of the benchmark are not comparable.
3.37
User experience · rank 4 of 5
- Unit
- score from 1 to 5, higher is better
- Configuration
- kimi-k2.6
- Measured
- 25 May 2026
- Not shown
- Not real customer satisfaction: the ratings come from a model-based judge.
57.1%
Average · rank 14 of 34
- Unit
- % resolved, higher is better
- Configuration
- kimi-k2.6
- Measured
- 30 Jun 2026
- Not shown
- Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
41.2%
Front end (SWE-Bench Multimodal) · rank 6 of 34
- Unit
- % resolved, higher is better
- Configuration
- kimi-k2.6
- Measured
- 30 Jun 2026
- Not shown
- Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
25.0%
Greenfield (Commit0) · rank 14 of 34
- Unit
- % resolved, higher is better
- Configuration
- kimi-k2.6
- Measured
- 30 Jun 2026
- Not shown
- Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
74.5%
Information gathering (GAIA) · rank 9 of 34
- Unit
- % resolved, higher is better
- Configuration
- kimi-k2.6
- Measured
- 30 Jun 2026
- Not shown
- Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
74.6%
Issue resolution (SWE-Bench) · rank 13 of 34
- Unit
- % resolved, higher is better
- Configuration
- kimi-k2.6
- Measured
- 30 Jun 2026
- Not shown
- Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
70.4%
Testing (SWT-Bench) · rank 12 of 34
- Unit
- % resolved, higher is better
- Configuration
- kimi-k2.6
- Measured
- 30 Jun 2026
- Not shown
- Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
34.0%
Requested-length error · rank 269 of 370
- Unit
- % off the requested word count, lower is better
- Configuration
- kimi-k2.6
- Measured
- 20 Apr 2026
- Not shown
- Not other format limits such as character counts or bullet counts.
18.0%
Requested-length error · rank 150 of 370
- Unit
- % off the requested word count, lower is better
- Configuration
- kimi-k2.6
- Measured
- 20 Apr 2026
- Not shown
- Not other format limits such as character counts or bullet counts.
0.34
Style adherence · rank 215 of 370
- Unit
- score from 0 to 1, higher is better
- Configuration
- kimi-k2.6
- Measured
- 20 Apr 2026
- Not shown
- Not brand voice on your own examples: UGI's prompts are private and lean towards creative writing.
0.35
Style adherence · rank 190 of 370
- Unit
- score from 0 to 1, higher is better
- Configuration
- kimi-k2.6
- Measured
- 20 Apr 2026
- Not shown
- Not brand voice on your own examples: UGI's prompts are private and lean towards creative writing.
64.2
Writing score · rank 72 of 370
- Unit
- score out of 100, higher is better
- Configuration
- kimi-k2.6
- Measured
- 20 Apr 2026
- Not shown
- Not business copy quality: UGI's prompts are private and lean towards creative writing, and models that often refuse get no score.
64.4
Writing score · rank 70 of 370
- Unit
- score out of 100, higher is better
- Configuration
- kimi-k2.6
- Measured
- 20 Apr 2026
- Not shown
- Not business copy quality: UGI's prompts are private and lean towards creative writing, and models that often refuse get no score.
99.7%
Answer rate · rank 44 of 108
- Unit
- % of documents, higher is better
- Configuration
- Kimi K2.6
- 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.
10.8%
Hallucination rate · rank 68 of 108
- Unit
- % of summaries, lower is better
- Configuration
- Kimi K2.6
- 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.
