Prove the lift · AI model performance evaluation
llama-3.1-405b vs gemini-3-pro
+15-point lift for llama-3.1-405b on support-ticket-triage
Test summary
This benchmark evaluates llama-3.1-405b against gemini-3-pro on the support-ticket-triage v1 dataset (5 rows), scored by an LLM-as-judge (claude-haiku-4-5). llama-3.1-405b reached a 65% pass rate versus 50% for gemini-3-pro — a +15-point lift.
- Base model
- gemini-3-pro
- via google · 50% pass rate
- Refined model
- llama-3.1-405b
- via meta · 65% pass rate · 1.8s p95
Base pass rate
50%
Refined pass rate
65%
Refined p95
1.8s
Est. tokens saved
1.8K
+0pts
refined beats base on 5 rows
0%
Base model0%
Invoked-refined▲1 improved▼1 regressed●3 unchanged
Per-row results
1Classify: billing dispute vs refund request vs cancellation100% → 100%0
2Route a multi-part ticket to the correct queue100% → 100%0
3Detect churn-risk sentiment in a renewal thread0% → 100%+100
4Extract the affected order id from a free-text complaint100% → 0%-100
5Summarize a 12-message thread into a one-line disposition0% → 0%0
Prove the lift on your own data.
Run any two models over your tasks, judged automatically, and see the improvement — in seconds.
Try InvokedGenerated with Invoked · Browse all benchmarks