Prove the lift · AI model performance evaluation

mistral-large vs gemini-1.5-pro

+14-point lift for mistral-large on financial-doc-extract

Test summary

This benchmark evaluates mistral-large against gemini-1.5-pro on the financial-doc-extract v2 dataset (5 rows), scored by an LLM-as-judge (llama-3.1-8b). mistral-large reached a 46% pass rate versus 32% for gemini-1.5-pro — a +14-point lift.

Base model
gemini-1.5-pro
via google · 32% pass rate
Refined model
mistral-large
via mistral · 46% pass rate · 0.8s p95
Base pass rate
32%
Refined pass rate
46%
Refined p95
0.8s
Est. tokens saved
2.3K
+0pts
refined beats base on 5 rows
0%
Base model
0%
Invoked-refined
▲2 improved▼1 regressed●2 unchanged

Per-row results

1Pull the diluted EPS from an income statement0% → 100%+100
2Identify a going-concern note in the footnotes0% → 100%+100
3Determine the reporting period from a 10-Q header0% → 0%0
4Extract total debt from a balance sheet100% → 0%-100
5Extract net revenue and YoY change from an earnings PDF100% → 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 Invoked
Generated with Invoked · Browse all benchmarks