Back to Archive

AutoBench Run 6 - August 2026

Twenty models scored on 252 questions across ten subject domains. Score, cost, and latency are the per-answer committee rank, billed cost, and generation time. Routers in the AutoBench Router Benchmark are scored against this answer matrix.

Latest
Date
August 29, 2026
Version
2026-08-29
Models
20
New Models
18

Run data

Model
Average (All Topics)Computer SciencePhilosophySocial ScienceLanguageScienceTechnologyArtsLiteratureHistoryAgentic
144.22s (#20)131.56s (#18)138.80s (#20)131.91s (#20)202.57s (#20)183.40s (#20)176.48s (#20)149.72s (#20)137.27s (#20)122.29s (#20)92.40s (#20)
93.06s (#19)199.94s (#20)75.55s (#19)78.23s (#19)126.02s (#19)78.12s (#15)144.21s (#19)101.77s (#19)64.07s (#19)57.07s (#19)46.17s (#19)
52.38s (#14)115.30s (#16)33.16s (#12)34.78s (#9)90.25s (#18)92.81s (#19)51.13s (#12)57.43s (#13)21.51s (#8)18.61s (#7)11.39s (#11)
66.32s (#18)146.26s (#19)36.84s (#13)75.42s (#18)79.54s (#17)75.50s (#14)89.32s (#18)75.40s (#17)55.66s (#18)39.16s (#18)10.18s (#8)
54.83s (#16)87.51s (#10)50.23s (#16)51.38s (#15)74.81s (#16)63.71s (#12)65.68s (#16)85.49s (#18)39.19s (#14)30.41s (#14)14.13s (#15)
53.47s (#15)124.68s (#17)51.88s (#18)49.07s (#14)64.52s (#15)82.12s (#16)58.67s (#13)52.95s (#12)29.77s (#9)26.00s (#12)12.05s (#13)
48.26s (#12)109.47s (#14)17.15s (#5)33.18s (#8)57.96s (#14)88.26s (#18)61.40s (#15)36.66s (#8)19.39s (#7)22.65s (#10)27.60s (#18)
56.52s (#17)110.46s (#15)48.75s (#15)73.73s (#17)55.36s (#13)55.20s (#11)87.72s (#17)62.84s (#15)52.06s (#17)30.99s (#15)16.39s (#16)
38.07s (#8)85.53s (#9)32.16s (#11)32.24s (#7)54.21s (#12)39.80s (#5)58.82s (#14)44.18s (#11)18.58s (#6)17.26s (#6)11.54s (#12)
45.07s (#11)104.13s (#12)40.08s (#14)44.04s (#12)52.21s (#11)84.02s (#17)45.25s (#9)29.84s (#7)30.73s (#11)21.52s (#9)9.20s (#4)
48.43s (#13)105.93s (#13)30.98s (#10)45.86s (#13)49.83s (#10)65.90s (#13)46.95s (#10)60.12s (#14)39.80s (#15)34.72s (#17)12.96s (#14)
44.20s (#10)93.03s (#11)50.94s (#17)42.12s (#10)40.56s (#9)44.14s (#8)48.83s (#11)70.62s (#16)33.08s (#12)32.76s (#16)10.99s (#10)
38.26s (#9)78.66s (#8)25.37s (#7)55.66s (#16)36.62s (#8)51.44s (#10)22.60s (#4)43.49s (#10)44.15s (#16)22.74s (#11)10.11s (#7)
32.29s (#6)47.02s (#6)29.06s (#8)42.45s (#11)31.53s (#7)40.14s (#6)43.00s (#8)38.71s (#9)29.99s (#10)21.01s (#8)9.72s (#6)
33.32s (#7)54.69s (#7)30.81s (#9)27.26s (#6)29.87s (#6)50.00s (#9)40.63s (#7)23.53s (#4)37.73s (#13)29.77s (#13)17.40s (#17)
26.58s (#5)46.53s (#5)24.12s (#6)25.19s (#5)27.94s (#5)41.42s (#7)32.97s (#6)27.54s (#5)15.57s (#4)16.91s (#5)10.65s (#9)
19.37s (#3)36.26s (#3)12.26s (#2)14.99s (#1)21.19s (#4)38.89s (#4)19.93s (#3)14.72s (#1)14.10s (#2)10.63s (#1)9.25s (#5)
21.56s (#4)40.61s (#4)16.87s (#4)20.71s (#4)21.16s (#3)27.47s (#2)27.44s (#5)27.79s (#6)16.56s (#5)15.89s (#4)6.09s (#2)
17.02s (#2)23.97s (#1)15.27s (#3)15.83s (#2)19.18s (#2)28.57s (#3)19.91s (#2)15.65s (#2)11.57s (#1)11.51s (#2)8.75s (#3)
15.89s (#1)29.76s (#2)8.75s (#1)17.08s (#3)13.16s (#1)27.07s (#1)18.45s (#1)17.17s (#3)14.35s (#3)11.54s (#3)3.13s (#1)
AutoBench Run 6 - August 2026 - AutoBench