Prove the lift · AI model performance evaluation

gpt-4.1 vs gpt-oss-120b

+9-point lift for gpt-4.1 on ecommerce-catalog-norm

Test summary

This benchmark evaluates gpt-4.1 against gpt-oss-120b on the ecommerce-catalog-norm v3 dataset (5 rows), scored by an LLM-as-judge (gpt-4o-mini). gpt-4.1 reached a 51% pass rate versus 42% for gpt-oss-120b — a +9-point lift.

Base model
gpt-oss-120b
via openai · 42% pass rate
Refined model
gpt-4.1
via openai · 51% pass rate · 0.7s p95
Base pass rate
42%
Refined pass rate
51%
Refined p95
0.7s
Est. tokens saved
11.0K
+0pts
refined beats base on 5 rows
0%
Base model
0%
Invoked-refined
0 improved0 regressed4 unchanged

Per-row results

1Normalize a free-text size into a standard enum0%0%0
2Deduplicate two near-identical SKUs0%0%0
3Classify a listing into the right category node100%100%0
4Extract material + color from a product title0%0%0
Prove the lift on your own data.

Run any two models over your tasks, judged automatically, and see the improvement — in seconds.

Try Invoked
Generated with Invoked · Browse all benchmarks