Qwen/Qwen3.5-397B-A17B
ko · 85.3 source-defined score; unit unspecified
Source-qualified evaluations, report statistics and safety observations.
ko · 85.3 source-defined score; unit unspecified
mmlu_pro · 84.4 source-defined score; unit unspecified
terminalbench_2 · 51 source-defined score; unit unspecified
swe_bench_%_resolved · 64.2 source-defined score; unit unspecified
overall · 90.72 source-defined score; unit unspecified
general · 58.2 source-defined score; unit unspecified
mcq_4_choices · 40.66 source-defined score; unit unspecified
swe_bench_multilingual_%_resolved · 72.4 source-defined score; unit unspecified
apex-agents · 14.4 source-defined score; unit unspecified
overall · 54.2 source-defined score; unit unspecified
toolathlon_verified · 72.5 source-defined score; unit unspecified
mmlu_pro · 37.9 source-defined score; unit unspecified
illustrator_windows · 19.4 source-defined score; unit unspecified
non_latin · 80.7 source-defined score; unit unspecified
OmniCompass-VL · 26.1 source-defined score; unit unspecified
deep_swe · 56 source-defined score; unit unspecified
skillsbench_v1_1 · 68.7 source-defined score; unit unspecified
hle · 26.7 source-defined score; unit unspecified
These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.