Prove the lift · AI model performance evaluation

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

-10-point lift for claude-opus-4-8 on medical-coding-extract

Test summary

This benchmark evaluates claude-opus-4-8 against claude-opus-4-6 on the medical-coding-extract v2 dataset (5 rows), scored by an LLM-as-judge (claude-haiku-4-5). claude-opus-4-8 reached a 45% pass rate versus 55% for claude-opus-4-6 — a -10-point lift.

Base model
claude-opus-4-6
via anthropic · 55% pass rate
Refined model
claude-opus-4-8
via anthropic · 45% pass rate · 1.9s p95
Base pass rate
55%
Refined pass rate
45%
Refined p95
1.9s
Est. tokens saved
0pts
no lift — the base model held its ground
0%
Base model
0%
Invoked-refined
2 improved1 regressed2 unchanged

Per-row results

1Determine the correct E/M level from a visit note100%0%-100
2Pull the procedure code from an operative report0%100%+100
3Extract medication + dosage from a clinical note0%0%0
4Map a discharge summary to the primary ICD-10 code0%100%+100
5Identify a contraindication from a problem list100%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