Prove the lift · AI model performance evaluation

claude-opus-4-6 vs claude-opus-4-8

-2-point lift for claude-opus-4-6 on contract-clause-qa

Test summary

This benchmark evaluates claude-opus-4-6 against claude-opus-4-8 on the contract-clause-qa v1 dataset (5 rows), scored by an LLM-as-judge (llama-3.1-8b). claude-opus-4-6 reached a 40% pass rate versus 42% for claude-opus-4-8 — a -2-point lift.

Base model
claude-opus-4-8
via anthropic · 42% pass rate
Refined model
claude-opus-4-6
via anthropic · 40% pass rate · 2.0s p95
Base pass rate
42%
Refined pass rate
40%
Refined p95
2.0s
Est. tokens saved
953
−0pts
no lift — the base model held its ground
0%
Base model
0%
Invoked-refined
▲1 improved▼0 regressed●4 unchanged

Per-row results

1Identify the governing-law clause and jurisdiction0% → 100%+100
2Extract the auto-renewal notice window in days100% → 100%0
3Find the liability cap and its exceptions0% → 0%0
4Determine indemnification scope and carve-outs0% → 0%0
5Locate the termination-for-convenience terms100% → 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