AIhoster
An independent Russian benchmark aggregation site that collects and ranks results from model cards, papers, and public leaderboard reports without re-running evaluations; each record is bound to an exact repository ID and evaluation configuration.
Timeline (last 90 days)
As of October 19, 2026, Darwin-180B-RSI ranks first on the AIhoster leaderboard for MMLU-Pro at 88.1% and first for GPQA Diamond at 94.4%.
AIhoster is an independent benchmark aggregation site whose practice is to collect scores reported in model cards, papers, and public leaderboards, without re-running benchmarks or building a normalized composite index.
Under the AIhoster methodology, Darwin-180B-RSI's scores come from VIDRAFT's own model card report and have not been independently reproduced by a third party.
All Facts (3)
As of October 19, 2026, Darwin-180B-RSI ranks first on the AIhoster leaderboard for MMLU-Pro at 88.1% and first for GPQA Diamond at 94.4%.
50%UnverifiedAIhoster is an independent benchmark aggregation site whose practice is to collect scores reported in model cards, papers, and public leaderboards, without re-running benchmarks or building a normalized composite index.
50%UnverifiedUnder the AIhoster methodology, Darwin-180B-RSI's scores come from VIDRAFT's own model card report and have not been independently reproduced by a third party.
50%