Prove the lift · AI model performance evaluation
grok-3 vs gpt-4.1-mini
-2-point lift for grok-3 on financial-doc-extract
Test summary
This benchmark evaluates grok-3 against gpt-4.1-mini on the financial-doc-extract v3 dataset (5 rows), scored by an LLM-as-judge (gpt-4o-mini, run locally). grok-3 reached a 45% pass rate versus 47% for gpt-4.1-mini — a -2-point lift.
- Base model
- gpt-4.1-mini
- via openai · 47% pass rate
- Refined model
- grok-3
- via xai · 45% pass rate · 1.7s p95
Base pass rate
47%
Refined pass rate
45%
Refined p95
1.7s
Est. tokens saved
219
−0pts
no lift — the base model held its ground
0%
Base model0%
Invoked-refined▲1 improved▼0 regressed●4 unchanged
Per-row results
1Extract net revenue and YoY change from an earnings PDF0% → 100%+100
2Determine the reporting period from a 10-Q header0% → 0%0
3Pull the diluted EPS from an income statement100% → 100%0
4Identify a going-concern note in the footnotes0% → 0%0
5Extract total debt from a balance sheet100% → 100%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