Skip to main content
PolicyBench

Gemini 3.5 Flash Lite

Google · AI alone, no tools
Current board, snapshot 2026-09-05
United States rank
#19 of 39
US exact
78.6%
United States

United States benchmark

Exact match
78.6%
Bounded score
88.3%
Parse rate
100.0%
Eligibility flags
934/979

Exact match by programall 18 output groups, best first

Bar length is the unweighted exact-match rate on that program's outputs; the right column is the program's share of the headline weight, which follows household dollars, so the bars do not average to the headline.

  1. Local income tax100.0%0.3%
  2. Person-level Medicare eligibility100.0%11%
  3. Reduced-price school meals eligibility99.0%0.1%
  4. Person-level WIC eligibility98.3%0.3%
  5. TANF98.0%0.3%
  6. SSI97.9%2.0%
  7. Person-level Head Start eligibility97.4%1.2%
  8. Free school meals eligibility96.0%0.7%
  9. Self-employment tax95.0%2.1%
  10. Person-level Early Head Start eligibility94.7%3.1%
  11. Person-level CHIP eligibility94.4%0.2%
  12. Person-level Medicaid eligibility86.4%30%
  13. Federal refundable credits86.0%3.7%
  14. SNAP82.5%4.1%
  15. State refundable credits79.0%0.6%
  16. Payroll tax78.0%16%
  17. State tax before refundable credits53.0%6.0%
  18. Federal tax before refundable credits49.0%19%
Table view (exact, within 1%, outputs), hardest first
ProgramExactWithin 1%n
Federal tax before refundable credits49.0%51.0%100
State tax before refundable credits53.0%53.0%100
Payroll tax78.0%82.0%100
State refundable credits79.0%79.0%100
SNAP82.5%82.5%97
Federal refundable credits86.0%86.0%100
Person-level Medicaid eligibility86.4%86.4%177
Person-level CHIP eligibility94.4%94.4%177
Person-level Early Head Start eligibility94.7%94.7%38
Self-employment tax95.0%97.0%100
Free school meals eligibility96.0%96.0%100
Person-level Head Start eligibility97.4%97.4%38
SSI97.9%97.9%97
TANF98.0%98.0%100
Person-level WIC eligibility98.3%98.3%177
Reduced-price school meals eligibility99.0%99.0%100
Local income tax100.0%100.0%100
Person-level Medicare eligibility100.0%100.0%172

Hardest casesworst misses on positive references

Scores are from the frozen manuscript snapshot under the AI-alone condition: structured responses over the same household facts (whole-scenario or in output subsets, per the serving-configuration table), no tools, graded against PolicyEngine reference outputs. See the leaderboard and paper for methodology. The frozen US annotations cover 8,783 rows whose legacy threshold score is below 1. That universe contains 8,780 of 8,780 exact-match misses and 3 exact hits. Another 1,605 rows with bounded score below 100 were not selected and have no audit annotation.