Prove the lift · AI model performance evaluation
gemini-3-pro vs gpt-5.1
+8-point lift for gemini-3-pro on ecommerce-catalog-norm
Test summary
This benchmark evaluates gemini-3-pro against gpt-5.1 on the ecommerce-catalog-norm v1 dataset (5 rows), scored by an LLM-as-judge (gpt-4o-mini). gemini-3-pro reached a 43% pass rate versus 35% for gpt-5.1 — a +8-point lift.
- Base model
- gpt-5.1
- via openai · 35% pass rate
- Refined model
- gemini-3-pro
- via google · 43% pass rate · 2.8s p95
Base pass rate
35%
Refined pass rate
43%
Refined p95
2.8s
Est. tokens saved
8.4K
+0pts
refined beats base on 5 rows
0%
Base model0%
Invoked-refined▲1 improved▼1 regressed●2 unchanged
Per-row results
1Extract material + color from a product title100% → 0%-100
2Deduplicate two near-identical SKUs100% → 100%0
3Normalize a free-text size into a standard enum0% → 100%+100
4Classify a listing into the right category node0% → 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 InvokedGenerated with Invoked · Browse all benchmarks