Prove the lift · AI model performance evaluation
gpt-5 vs glm-5.2
+16-point lift for gpt-5 on support-ticket-triage
Test summary
This benchmark evaluates gpt-5 against glm-5.2 on the support-ticket-triage v1 dataset (6 rows), scored by an LLM-as-judge (claude-haiku-4-5). gpt-5 reached a 65% pass rate versus 49% for glm-5.2 — a +16-point lift.
- Base model
- glm-5.2
- via z-ai · 49% pass rate
- Refined model
- gpt-5
- via openai · 65% pass rate · 0.9s p95
Base pass rate
49%
Refined pass rate
65%
Refined p95
0.9s
Est. tokens saved
5.1K
+0pts
refined beats base on 6 rows
0%
Base model0%
Invoked-refined▲2 improved▼1 regressed●3 unchanged
Per-row results
1Summarize a 12-message thread into a one-line disposition100% → 100%0
2Priority-tag an outage report vs a feature question0% → 100%+100
3Extract the affected order id from a free-text complaint100% → 100%0
4Detect churn-risk sentiment in a renewal thread0% → 100%+100
5Route a multi-part ticket to the correct queue100% → 100%0
6Classify: billing dispute vs refund request vs cancellation100% → 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 InvokedGenerated with Invoked · Browse all benchmarks