BlogLLM Benchmark Scores as Published, Side by Side

WinoGrande: Results as Published

Every result the catalog holds for WinoGrande: 6 models from 3 developers, read from 2 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.

Area General Knowledge and ReasoningMaintainer Allen Institute for AI (Keisuke Sakaguchi and co-authors)Introduced 2019Data version 2026103001, as of 2026-10-30

WinoGrande tests commonsense reasoning with 44,000 pronoun-resolution problems in the style of the Winograd Schema Challenge, built at scale and filtered to remove shortcuts. The score is the share of problems answered correctly.

WinoGrande

ModelValuePrinted bySetup as printed
DeepSeek-V4-Flash-BaseDeepSeek, released 2026-04-2479.5%DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15; printed on 1 more page-
DeepSeek-V4-Pro-BaseDeepSeek, released 2026-04-2481.5%DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15; printed on 1 more page-
DeepSeek-V3.2-BaseDeepSeek, released 2025-12-0178.9%DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15-
Kimi K2 BaseMoonshot AI, released 2025-07-1185.3%Xiaomi, Xiaomi MiMo-V2.5-Pro model cardchecked 2026-10-15Harness: 5-shot
MiMo-V2.5 BaseXiaomi, released 2026-04-2784.7%Xiaomi, Xiaomi MiMo-V2.5-Pro model cardchecked 2026-10-15Harness: 5-shot
MiMo-V2.5-Pro BaseXiaomi, released 2026-04-2785.6%Xiaomi, Xiaomi MiMo-V2.5-Pro model cardchecked 2026-10-15Harness: 5-shot

Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.