Prove the lift · AI model performance evaluation

gpt-4-turbo vs mistral-large

-9-point lift for gpt-4-turbo on insurance-extraction-eval

Test summary

This benchmark evaluates gpt-4-turbo against mistral-large on the insurance-extraction-eval v1 dataset (7 rows), scored by an LLM-as-judge (llama-3.1-8b). gpt-4-turbo reached a 32% pass rate versus 41% for mistral-large — a -9-point lift.

Base model
mistral-large
via mistral · 41% pass rate
Refined model
gpt-4-turbo
via openai · 32% pass rate · 1.9s p95
Base pass rate
41%
Refined pass rate
32%
Refined p95
1.9s
Est. tokens saved
267
0pts
no lift — the base model held its ground
0%
Base model
0%
Invoked-refined
2 improved0 regressed5 unchanged

Per-row results

1Subrogation waiver interaction with additional-insured status100%100%0
2Coinsurance 80% clause recovery on an under-insured building100%100%0
3Business-interruption period-of-restoration calculation0%0%0
4Suit-limitation clause: is a late-filed claim contractually barred?0%100%+100
5Replacement cost vs ACV on a depreciated roof0%0%0
6Aggregate deductible across a loss sequence under a per-occurrence + annual cap0%0%0
7Named-peril coverage determination under a basic HO form0%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 Invoked
Generated with Invoked · Browse all benchmarks