Claude 1 and gpt-oss-20b don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Claude 1 shipped 875 days before gpt-oss-20b, so benchmark comparisons should account for the intervening progress.

Claude 1 is proprietary, while gpt-oss-20b is open weight.

Direct benchmark comparisons are unavailable — Claude 1 and gpt-oss-20b don't publish scores on any of the same benchmarks.