Card-wide default max effort; one agent; 166 retained golden tasks; five sequential episodes, each with a fresh context budget up to 1M tokens; no internet or decompilation tools.
未知项
未另记
工具
工具条件未记录
推理预算
max
框架
未记录
协议
ProgramBench; 166 golden tasks; hidden-test pass rate after episode 1.
条件
Single-agent; episode 1; each episode has a fresh context budget up to 1M tokens; max effort by card-wide default.
Card-wide default max effort; one agent; 166 retained golden tasks; five sequential episodes, each with a fresh context budget up to 1M tokens; no internet or decompilation tools.
未知项
未另记
工具
工具条件未记录
推理预算
max
框架
未记录
协议
ProgramBench; 166 golden tasks; hidden-test pass rate after episode 5.
条件
Single-agent; five sequential episodes continue from the previous codebase; fresh context budget up to 1M tokens per episode; max effort by card-wide default.
System Card max effort (per-result benchmark harnesses)
设置
System Card §8.1 default: adaptive thinking/max effort and default sampling settings unless a section says otherwise. Benchmark tools, evaluator, safeguards, fallback, and repetitions remain per-result conditions; this config represents the stated model effort, not harness equivalence.
未知项
Exact run time is not disclosed.、Per-benchmark tools, safeguards, and repetition counts are specified only where the source states them.
工具
工具条件未记录
推理预算
max
框架
mini-SWE-agent harness
协议
Single-agent ProgramBench; 166 retained golden tasks; mini-swe-agent harness; no six-hour time limit; no internet or decompilation.
条件
The four-task difference from the multi-agent section is not merged; score uses only tests passed by each reference binary. Attempt count is not stated. Source-wide conditions: No internet or decompilation; score only tests passed by the reference binary; do not apply the 91.2% to all 200 tasks.
Claude Opus · ProgramBench · 166 个保留的 golden tasks;架构与时限按协议区分 · Release history
View scores, configurations, and disclosures for ProgramBench · 166 个保留的 golden tasks;架构与时限按协议区分 across 2 Claude Opus releases, ordered by release date.
Differences are calculated only for unique, active, exact values between direct predecessors when the unit, metric, and explicitly recorded configuration and conditions match.
· 2 score records for this release, including 2 exact values (may include historical records)
Claude Opus 5 · ProgramBench scores and configurations
Metric and score
Configuration and evaluation conditions
Sources
Hidden test pass rate83%Reported 83% · Unit %
Published Unconfirmed
ProgramBench · single agent
Settings
Card-wide default max effort; one agent; 166 retained golden tasks; five sequential episodes, each with a fresh context budget up to 1M tokens; no internet or decompilation tools.
Unknowns
none recorded
Tools
Tool access not recorded
Reasoning budget
max
Framework
not recorded
Protocol
ProgramBench; 166 golden tasks; hidden-test pass rate after episode 1.
Conditions
Single-agent; episode 1; each episode has a fresh context budget up to 1M tokens; max effort by card-wide default.
Card-wide default max effort; one agent; 166 retained golden tasks; five sequential episodes, each with a fresh context budget up to 1M tokens; no internet or decompilation tools.
Unknowns
none recorded
Tools
Tool access not recorded
Reasoning budget
max
Framework
not recorded
Protocol
ProgramBench; 166 golden tasks; hidden-test pass rate after episode 5.
Conditions
Single-agent; five sequential episodes continue from the previous codebase; fresh context budget up to 1M tokens per episode; max effort by card-wide default.
· 1 score record for this release, including 1 exact value (may include historical records)
Records for the direct predecessor do not provide matching conditions for a valid difference.
Claude Opus 5.5 · ProgramBench scores and configurations
Metric and score
Configuration and evaluation conditions
Sources
Hidden test pass rate91.2%Reported 91.2% · Unit %
Published Unconfirmed
System Card max effort (per-result benchmark harnesses)
Settings
System Card §8.1 default: adaptive thinking/max effort and default sampling settings unless a section says otherwise. Benchmark tools, evaluator, safeguards, fallback, and repetitions remain per-result conditions; this config represents the stated model effort, not harness equivalence.
Unknowns
Exact run time is not disclosed.、Per-benchmark tools, safeguards, and repetition counts are specified only where the source states them.
Tools
Tool access not recorded
Reasoning budget
max
Framework
mini-SWE-agent harness
Protocol
Single-agent ProgramBench; 166 retained golden tasks; mini-swe-agent harness; no six-hour time limit; no internet or decompilation.
Conditions
The four-task difference from the multi-agent section is not merged; score uses only tests passed by each reference binary. Attempt count is not stated. Source-wide conditions: No internet or decompilation; score only tests passed by the reference binary; do not apply the 91.2% to all 200 tasks.