Skip to content

qwen3.5-397b-a17b

Alibaba

average±4.1 (95%): 67.7
leaderboard rankof 114: ≈70
benchmarksof 12: 10
per 1,000 documentsmean over its benchmarks: $7.42
seconds per inputmean response time: 45.5

Compare with other models Its row in the Matrix All its test runs

By task

Data correction

59.0

Best of 114 models with all its benchmarks: claude-opus-5-5, 82.7.
Transcription

78.4

Best of 114 models with all its benchmarks: gemini-4-argon, 96.2.

By benchmark

Benchmark	Score	Rank	Per 1,000 docs	s / input	Best run
Bibliographic Data	97.5±2.0	≈8of 115	$9.35$0.047 run	37.6	T0866 · 25 Mar 2026prompt.txt
Blacklist Cards	90.2±3.6	≈51of 114	$2.36$0.078 run	8.27	T0873 · 25 Mar 2026prompt.txt
Book Advert XML Repair	21.3±6.9	≈64of 119	$32$1.62 run	296.25	T0876 · 25 Mar 2026prompt.txt
Business Letters	60.0±6.5	≈24of 114	$3.29$0.19 run	9.78	T0867 · 25 Mar 2026prompt.txt
Company Listsother prompt	84.3±7.9	≈29of 114	$5.26$0.079 run	19.91	T0875 · 25 Mar 2026prompt_min.txt
Fraktur Adverts	68.2±35.6	≈67of 114	$8.67$0.043 run	35.12	T0870 · 25 Mar 2026prompt_optimized.txt
Library Cards	88.6±1.4	≈15of 114	$1.73$0.45 run	6.46	T0871 · 25 Mar 2026prompt.txt
Magazine Pages	0.0±0.0	≈79of 115	$1.91$0.088 run	7.9	T0878 · 25 Mar 2026prompt.txt
Medieval Manuscripts	70.2±16.5	≈27of 114	$2.91$0.035 run	12.4	T0872 · 25 Mar 2026prompt.txt
Personnel Cards	96.7±2.5	≈30of 114	$6.21$0.38 run	21.34	T0877 · 25 Mar 2026prompt.txt

Not run on: Duty Rosters, General Meeting Minutes.

Price

From	Input, $ per million tokens	Output, $ per million tokens	Source
24 Mar 2026	0.6	3.6	archived price page

Every run