Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations) · 附件证据(ev-h55-card-healthbench-pro)
Claude Haiku 5.5 的 HealthBench Professional · 525 physician-authored conversations 官方出处。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。证据 ev-h55-card-healthbench-pro。
Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations),Healthcare evaluations / low配置,Length-adjusted score 57.9%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations),Healthcare evaluations / medium配置,Length-adjusted score 59.9%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations),Healthcare evaluations / high配置,Length-adjusted score 61.3%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations),Healthcare evaluations / xhigh配置,Length-adjusted score 61%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations),Healthcare evaluations / max配置,Length-adjusted score 64.8%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations),Healthcare evaluations / low配置,Raw score 61.6%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · HealthBench Professional(525 physician-authored conversations),Healthcare evaluations / max配置,Raw score 71%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A。 查看成绩条件
HealthBench Professional length-adjusted scores are labeled 57.9%, 59.9%, 61.3%, 61.0%, and 64.8%; the card also reports raw values of 61.6% at low and 71.0% at max.
Safety classifiers enabled; low effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
禁止工具
推理预算
low
框架
未记录
协议
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
条件
low effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; medium effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
禁止工具
推理预算
medium
框架
未记录
协议
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
条件
medium effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; high effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
禁止工具
推理预算
high
框架
未记录
协议
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
条件
high effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; xhigh effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
禁止工具
推理预算
xhigh
框架
未记录
协议
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
条件
xhigh effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; max effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
禁止工具
推理预算
max
框架
未记录
协议
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
条件
max effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations) · Attachment evidence (ev-h55-card-healthbench-pro)
Official source for HealthBench Professional · 525 physician-authored conversations in Claude Haiku 5.5. See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. Evidence ev-h55-card-healthbench-pro.
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / low, Length-adjusted score 57.9% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / medium, Length-adjusted score 59.9% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / high, Length-adjusted score 61.3% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / xhigh, Length-adjusted score 61% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / max, Length-adjusted score 64.8% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / low, Raw score 61.6% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / max, Raw score 71% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A. View score conditions
HealthBench Professional length-adjusted scores are labeled 57.9%, 59.9%, 61.3%, 61.0%, and 64.8%; the card also reports raw values of 61.6% at low and 71.0% at max.
Attachment · System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A
Material version
144 页 PDF/TXT:本地留存副本、current-haiku-5-5-system-card.txt;封面日期 2026-10-07;正文 pp.1-144 已全文核读,全部63个含图/向量图表页实际视觉检查(本轮补看53页,合并此前10页),无未读页或未查看图表页;此为 current 卡片,不作为首发日附件证据。
Version published
Disclosure date
Unconfirmed · Launch or later disclosure unconfirmed
Collected
Historical basis
No reliable historical version evidence available
Citation text
Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / low, Length-adjusted score 57.9% (exact score publication date unconfirmed); Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / medium, Length-adjusted score 59.9% (exact score publication date unconfirmed); Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / high, Length-adjusted score 61.3% (exact score publication date unconfirmed); Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / xhigh, Length-adjusted score 61% (exact score publication date unconfirmed); Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / max, Length-adjusted score 64.8% (exact score publication date unconfirmed); Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / low, Raw score 61.6% (exact score publication date unconfirmed); Claude Haiku 5.5 · HealthBench Professional (525 physician-authored conversations), configuration Healthcare evaluations / max, Raw score 71% (exact score publication date unconfirmed). Official source: Claude Haiku 5.5 System Card (Attachment, 144 页 PDF/TXT:本地留存副本、current-haiku-5-5-system-card.txt;封面日期 2026-10-07;正文 pp.1-144 已全文核读,全部63个含图/向量图表页实际视觉检查(本轮补看53页,合并此前10页),无未读页或未查看图表页;此为 current 卡片,不作为首发日附件证据。), System Card PDF pp.111, 133-134, Table 8.1.A and Figure 8.11.3.A, https://www.anthropic.com/claude-haiku-5-5-system-card#page=134. Benchlog https://benchlog.lekshome.cc/evidence/claude-haiku-5-5/ev-h55-card-healthbench-pro/?lang=en
Related scores and conditions
Claude Haiku 5.5 · HealthBench Professional scores and configurations
Metric and score
Configuration and evaluation conditions
Evidence relation
Length-adjusted score57.9%Reported 57.9% · Unit %
Published Unconfirmed
Healthcare evaluations / low
Settings
Safety classifiers enabled; low effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tools forbidden
Reasoning budget
low
Framework
not recorded
Protocol
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
Conditions
low effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; medium effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tools forbidden
Reasoning budget
medium
Framework
not recorded
Protocol
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
Conditions
medium effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; high effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tools forbidden
Reasoning budget
high
Framework
not recorded
Protocol
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
Conditions
high effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; xhigh effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tools forbidden
Reasoning budget
xhigh
Framework
not recorded
Protocol
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
Conditions
xhigh effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.
Safety classifiers enabled; max effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tools forbidden
Reasoning budget
max
Framework
not recorded
Protocol
HealthBench Professional length-adjusted score; Claude Opus 4.8 grader; no tools or customized system prompt; five runs per effort level.
Conditions
max effort. Low through xhigh runs were made October 4; max runs were made October 1-2. The card also reports 13 max-effort repeat runs on October 4 averaging 64.2%; kept separate from the primary 64.8% result. Labeled figure value.