Home Model series Claude Opus Claude Opus 4.8 Claude Opus 4.8 · BrowseComp · 1,266 题;单代理与多代理按协议区分 · Scores and configurations Claude Opus 4.8 · BrowseComp · 1,266 题;单代理与多代理按协议区分 · Scores and configurations Claude Opus 4.8 record for BrowseComp · 1,266 题;单代理与多代理按协议区分. 1 score record for this release, including 1 exact value (may include historical records). Includes 1 source location.
Scores and evaluation conditions Claude Opus 4.8 · BrowseComp scores and configurations Metric and score Configuration and evaluation conditions Sources Accuracy 84.3% Reported 84.3% · Unit % Published Unconfirmed
BrowseComp 10M Settings Web search、web fetch、programmatic tool calling、code execution;adaptive thinking max effort;10M token 总预算,200K 触发 context compaction;使用更新的污染规避 blocklist。 Unknowns none recorded Tools Tools allowed Reasoning budget max Framework not recorded Protocol Single-agent BrowseComp; updated contamination blocklist; search-result and fetched-page content filter also described in Appendix 9.3. Conditions Web search/fetch, programmatic tool calling and code execution; adaptive thinking/max effort; 10M token limit, compaction triggered at 200K. Exclude separate multi-agent result. View this score in rankings
Benchmark versions and metrics Aliases BrowseComp、Multi-agent BrowseComp
Accuracy · % · Higher is better