Skip to main content
PolicyBench

GPT-5.6 Luna

OpenAI · AI alone, no tools
United States rank
#2 of 23
US exact
84.5%
United States

United States benchmark

Exact match
84.5%
Bounded score
92.5%
Parse rate
100.0%
Eligibility flags
957/984

Score by programhardest first

ProgramExactWithin 1%n
State tax before refundable credits53.0%56.0%100
Federal tax before refundable credits65.0%70.0%100
SNAP78.0%81.0%100
State refundable credits79.0%79.0%100
Payroll tax89.0%91.0%100
Federal refundable credits90.0%90.0%100
Person-level Medicaid eligibility93.8%93.8%177
Person-level CHIP eligibility96.6%96.6%177
Reduced-price school meals eligibility97.0%97.0%100
SSI97.0%98.0%100
Person-level Medicare eligibility97.2%97.2%177
Free school meals eligibility98.0%98.0%100
Self-employment tax98.0%99.0%100
TANF99.0%99.0%100
Local income tax100.0%100.0%100
Person-level Early Head Start eligibility100.0%100.0%38
Person-level Head Start eligibility100.0%100.0%38
Person-level WIC eligibility100.0%100.0%177

Hardest casesworst misses on positive references

Scores are from the frozen manuscript snapshot under the AI-alone condition: one structured response per household, no tools, graded against PolicyEngine reference outputs. See the leaderboard and paper for methodology.