Prove the lift · AI model performance evaluation

gemini-2.5-pro vs qwen2.5-vl-32b

+27-point lift for gemini-2.5-pro on contract-clause-qa

Test summary

This benchmark evaluates gemini-2.5-pro against qwen2.5-vl-32b on the contract-clause-qa v3 dataset (5 rows), scored by an LLM-as-judge (claude-haiku-4-5). gemini-2.5-pro reached a 67% pass rate versus 40% for qwen2.5-vl-32b — a +27-point lift.

Base model
qwen2.5-vl-32b
via qwen · 40% pass rate
Refined model
gemini-2.5-pro
via google · 67% pass rate · 0.8s p95
Base pass rate
40%
Refined pass rate
67%
Refined p95
0.8s
Est. tokens saved
1.6K
+0pts
refined beats base on 5 rows
0%
Base model
0%
Invoked-refined
▲3 improved▼1 regressed●1 unchanged

Per-row results

1Find the liability cap and its exceptions0% → 100%+100
2Determine indemnification scope and carve-outs0% → 100%+100
3Extract the auto-renewal notice window in days0% → 100%+100
4Identify the governing-law clause and jurisdiction100% → 100%0
5Locate the termination-for-convenience terms100% → 0%-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