Prove the lift · AI model performance evaluation

minimax-m3 vs gpt-oss-120b

+11-point lift for minimax-m3 on insurance-extraction-eval

Test summary

This benchmark evaluates minimax-m3 against gpt-oss-120b on the insurance-extraction-eval v2 dataset (5 rows), scored by an LLM-as-judge (claude-haiku-4-5). minimax-m3 reached a 57% pass rate versus 46% for gpt-oss-120b — a +11-point lift.

Base model
gpt-oss-120b
via openai · 46% pass rate
Refined model
minimax-m3
via minimax · 57% pass rate · 3.3s p95
Base pass rate
46%
Refined pass rate
57%
Refined p95
3.3s
Est. tokens saved
1.8K
+0pts
refined beats base on 5 rows
0%
Base model
0%
Invoked-refined
▲3 improved▼1 regressed●1 unchanged

Per-row results

1Subrogation waiver interaction with additional-insured status100% → 0%-100
2Aggregate deductible across a loss sequence under a per-occurrence + annual cap0% → 100%+100
3Coinsurance 80% clause recovery on an under-insured building0% → 100%+100
4Suit-limitation clause: is a late-filed claim contractually barred?0% → 100%+100
5Business-interruption period-of-restoration calculation100% → 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