Prove the lift · AI model performance evaluation
gpt-4.1-mini vs gpt-oss-20b
+21-point lift for gpt-4.1-mini on medical-coding-extract
Test summary
This benchmark evaluates gpt-4.1-mini against gpt-oss-20b on the medical-coding-extract v2 dataset (5 rows), scored by an LLM-as-judge (gpt-4o-mini). gpt-4.1-mini reached a 58% pass rate versus 37% for gpt-oss-20b — a +21-point lift.
- Base model
- gpt-oss-20b
- via openai · 37% pass rate
- Refined model
- gpt-4.1-mini
- via openai · 58% pass rate · 1.9s p95
Base pass rate
37%
Refined pass rate
58%
Refined p95
1.9s
Est. tokens saved
2.3K
+0pts
refined beats base on 5 rows
0%
Base model0%
Invoked-refined▲2 improved▼0 regressed●3 unchanged
Per-row results
1Map a discharge summary to the primary ICD-10 code100% → 100%0
2Extract medication + dosage from a clinical note100% → 100%0
3Pull the procedure code from an operative report100% → 100%0
4Determine the correct E/M level from a visit note0% → 100%+100
5Identify a contraindication from a problem list0% → 100%+100
Prove the lift on your own data.
Run any two models over your tasks, judged automatically, and see the improvement — in seconds.
Try InvokedGenerated with Invoked · Browse all benchmarks