Claude Sonnet 5 · BrowseComp(1,266 题;单代理与多代理按协议区分) · 附件证据(ev5-browsecomp-scaling-chart)
Claude Sonnet 5 的 BrowseComp · 1,266 题;单代理与多代理按协议区分 官方出处。见《Claude Sonnet 5 System Card》§8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels。证据 ev5-browsecomp-scaling-chart。
Claude Sonnet 5 · BrowseComp(1,266 题;单代理与多代理按协议区分),BrowseComp max effort, 1M-token limit配置,Accuracy 79.3%(具体成绩公布时间未确认)。见《Claude Sonnet 5 System Card》§8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels。 查看成绩条件
Claude Sonnet 5 · BrowseComp(1,266 题;单代理与多代理按协议区分),BrowseComp max effort, 3M-token limit配置,Accuracy 82.9%(具体成绩公布时间未确认)。见《Claude Sonnet 5 System Card》§8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels。 查看成绩条件
附件 · §8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels
材料版本
PDF with changelog dated 2026-07-10; 146 pages; collected 2026-11-10
版本公布
未确认
披露时间
未确认 · 首发/后补未确认
采集时间
历史依据
暂无可靠历史版本依据
引用文本
Claude Sonnet 5 · BrowseComp(1,266 题;单代理与多代理按协议区分),BrowseComp max effort, 1M-token limit配置,Accuracy 79.3%(具体成绩公布时间未确认);Claude Sonnet 5 · BrowseComp(1,266 题;单代理与多代理按协议区分),BrowseComp max effort, 3M-token limit配置,Accuracy 82.9%(具体成绩公布时间未确认)。官方出处 Claude Sonnet 5 System Card(附件,PDF with changelog dated 2026-07-10; 146 pages; collected 2026-11-10)§8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels,https://www.anthropic.com/claude-sonnet-5-system-card#page=124。Benchlog https://benchlog.lekshome.cc/evidence/claude-sonnet-5/ev5-browsecomp-scaling-chart/
关联成绩与条件
Claude Sonnet 5 · BrowseComp 成绩与配置
指标与成绩
配置与评测条件
证据关系
Accuracy79.3%原文 79.3% · 单位 %
公布 未确认
BrowseComp max effort, 1M-token limit
设置
Adaptive thinking at max effort with web search, web fetch, programmatic tool calling, and code execution; 1M-token limit.
未知项
Whether context compaction was used at this token limit is not stated.
工具
允许工具
推理预算
max
框架
未记录
协议
BrowseComp Figure 8.9.2.A, test-time compute scaling at max effort with web search, web fetch, programmatic tool calling, and code execution.
条件
1M-token limit as labeled. The figure gives no exact per-point cost label.
直接支持该成绩
Accuracy82.9%原文 82.9% · 单位 %
公布 未确认
BrowseComp max effort, 3M-token limit
设置
Adaptive thinking at max effort with web search, web fetch, programmatic tool calling, and code execution; 3M-token limit; context compaction is reported as triggering at 200K tokens when extending beyond the 1M context window.
未知项
Sampling parameters are not stated.
工具
允许工具
推理预算
max
框架
未记录
协议
BrowseComp Figure 8.9.2.A, test-time compute scaling at max effort with web search, web fetch, programmatic tool calling, and code execution.
条件
3M-token limit as labeled; §8.9.2 reports context compaction triggered at 200K tokens to extend beyond the 1M context window. The figure gives no exact per-point cost label.
Official source for BrowseComp · 1,266 题;单代理与多代理按协议区分 in Claude Sonnet 5. See “Claude Sonnet 5 System Card” at §8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels. Evidence ev5-browsecomp-scaling-chart.
Claude Sonnet 5 · BrowseComp (1,266 题;单代理与多代理按协议区分), configuration BrowseComp max effort, 1M-token limit, Accuracy 79.3% (exact score publication date unconfirmed). See “Claude Sonnet 5 System Card” at §8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels. View score conditions
Claude Sonnet 5 · BrowseComp (1,266 题;单代理与多代理按协议区分), configuration BrowseComp max effort, 3M-token limit, Accuracy 82.9% (exact score publication date unconfirmed). See “Claude Sonnet 5 System Card” at §8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels. View score conditions
Attachment · §8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels
Material version
PDF with changelog dated 2026-07-10; 146 pages; collected 2026-11-10
Version published
Unconfirmed
Disclosure date
Unconfirmed · Launch or later disclosure unconfirmed
Collected
Historical basis
No reliable historical version evidence available
Citation text
Claude Sonnet 5 · BrowseComp (1,266 题;单代理与多代理按协议区分), configuration BrowseComp max effort, 1M-token limit, Accuracy 79.3% (exact score publication date unconfirmed); Claude Sonnet 5 · BrowseComp (1,266 题;单代理与多代理按协议区分), configuration BrowseComp max effort, 3M-token limit, Accuracy 82.9% (exact score publication date unconfirmed). Official source: Claude Sonnet 5 System Card (Attachment, PDF with changelog dated 2026-07-10; 146 pages; collected 2026-11-10), §8.9.2 Figure 8.9.2.A, PDF p.124, orange Claude Sonnet 5 series labels, https://www.anthropic.com/claude-sonnet-5-system-card#page=124. Benchlog https://benchlog.lekshome.cc/evidence/claude-sonnet-5/ev5-browsecomp-scaling-chart/?lang=en
Related scores and conditions
Claude Sonnet 5 · BrowseComp scores and configurations
Metric and score
Configuration and evaluation conditions
Evidence relation
Accuracy79.3%Reported 79.3% · Unit %
Published Unconfirmed
BrowseComp max effort, 1M-token limit
Settings
Adaptive thinking at max effort with web search, web fetch, programmatic tool calling, and code execution; 1M-token limit.
Unknowns
Whether context compaction was used at this token limit is not stated.
Tools
Tools allowed
Reasoning budget
max
Framework
not recorded
Protocol
BrowseComp Figure 8.9.2.A, test-time compute scaling at max effort with web search, web fetch, programmatic tool calling, and code execution.
Conditions
1M-token limit as labeled. The figure gives no exact per-point cost label.
Directly supports this score
Accuracy82.9%Reported 82.9% · Unit %
Published Unconfirmed
BrowseComp max effort, 3M-token limit
Settings
Adaptive thinking at max effort with web search, web fetch, programmatic tool calling, and code execution; 3M-token limit; context compaction is reported as triggering at 200K tokens when extending beyond the 1M context window.
Unknowns
Sampling parameters are not stated.
Tools
Tools allowed
Reasoning budget
max
Framework
not recorded
Protocol
BrowseComp Figure 8.9.2.A, test-time compute scaling at max effort with web search, web fetch, programmatic tool calling, and code execution.
Conditions
3M-token limit as labeled; §8.9.2 reports context compaction triggered at 200K tokens to extend beyond the 1M context window. The figure gives no exact per-point cost label.