Verdict
Quality matched on 100% of cases and the candidate is cheaper. Latency is informational during calibration.
Cases evaluated
Evaluation complete
Quality matches
100%
Est. monthly savings
Projected from matching events observed over 30 days
Status
The experiment finished and its verdict is available.
Controlled replay compares both models from the same worker. Production latency is shown separately as context and is not subtracted from candidate replay latency.
claude-sonnet-4-6 and claude-haiku-4-5 received the same frozen prompts in alternating order.
Replayed source p50
1,419 ms
claude-sonnet-4-6
Replayed candidate p50
626 ms
claude-haiku-4-5
Paired median delta
-762 ms (-55%)
95% interval: -874 ms to -690 ms
Controlled p90
1,624 ms / 735 ms
Source / candidate
Production source context
1,027 ms p50, 2,905 ms p90
30 captured measurements from the user's environment
Controlled coverage
30 successful pairs
Only complete source/candidate pairs enter the comparison (provider-completion-v1)
The evaluation compares structured content canonically and reports presentation adherence as separate evidence unless strict formatting is enabled for the feature.
This run's encrypted evidence expires at the date shown above, 14 days after the run was created. The verdict summary and savings estimate remain in experiment history after that run-specific evidence is deleted.
Review the case results before opening sensitive prompt and response evidence. Opening a case is audited.
| Case | Result | Format | Evaluation confidence | Attempts | Production source | Replayed source | Candidate | Replay delta | Status | Case details |
|---|---|---|---|---|---|---|---|---|---|---|
| 11 | Equivalent | Met | 97% | 1 | 4,703 ms 216,723 in / 9,382 out | 1,670 ms 216,723 in / 9,382 out | 560 ms 208,054 in / 7,881 out | -1,110 ms (-66%) | Evaluated |
Replay failures
0 source, 0 candidate
Failures are excluded from percentiles and retained as reliability evidence
| View details |
| 12 | Equivalent | Met | 97% | 2 | 577 ms 62,275 in / 3,002 out | 1,207 ms 62,275 in / 3,002 out | 591 ms 59,784 in / 2,522 out | -616 ms (-51%) | Evaluated | View details |
| 13 | Equivalent | Met | 97% | 1 | 1,119 ms 181,340 in / 9,829 out | 1,254 ms 181,340 in / 9,829 out | 622 ms 174,086 in / 8,256 out | -632 ms (-50%) | Evaluated | View details |
| 14 | Equivalent | Met | 97% | 1 | 1,141 ms 41,575 in / 1,698 out | 1,301 ms 41,575 in / 1,698 out | 653 ms 39,912 in / 1,426 out | -648 ms (-50%) | Evaluated | View details |
| 15 | Equivalent | Met | 97% | 1 | 365 ms 96,597 in / 4,414 out | 1,348 ms 96,597 in / 4,414 out | 684 ms 92,733 in / 3,708 out | -664 ms (-49%) | Evaluated | View details |
| 16 | Equivalent | Met | 97% | 1 | 2,532 ms 212,300 in / 8,488 out | 1,395 ms 212,300 in / 8,488 out | 715 ms 203,808 in / 7,130 out | -680 ms (-49%) | Evaluated | View details |
| 17 | Equivalent | Met | 97% | 1 | 1,897 ms 60,859 in / 2,716 out | 1,442 ms 60,859 in / 2,716 out | 746 ms 58,425 in / 2,281 out | -696 ms (-48%) | Evaluated | View details |
| 18 | Equivalent | Met | 97% | 1 | 204 ms 26,007 in / 1,019 out | 1,489 ms 26,007 in / 1,019 out | 517 ms 24,967 in / 856 out | -972 ms (-65%) | Evaluated | View details |
| 19 | Equivalent | Met | 97% | 1 | 361 ms 65,105 in / 2,859 out | 1,536 ms 65,105 in / 2,859 out | 548 ms 62,501 in / 2,402 out | -988 ms (-64%) | Evaluated | View details |
| 20 | Equivalent | Met | 97% | 1 | 2,124 ms 94,297 in / 4,995 out | 1,583 ms 94,297 in / 4,995 out | 579 ms 90,525 in / 4,196 out | -1,004 ms (-63%) | Evaluated | Hide details |
Case 20 evidenceFull captured evidence for this scored comparison. EquivalentFormat metscored97% evaluation confidenceJudge: gemini-3.1-pro-preview Evaluation notesThe candidate preserved the original answer's policy, resolution, and customer-safe guidance. Format evidenceThe response preserved the expected support-answer structure. Captured prompt | ||||||||||
Northstar support case 20: resolve the ticket-summary request with the account policy and cited help-center context.
claude-sonnet-4-6
Resolved ticket-summary case 20 with the correct policy, next step, and customer-safe explanation.
claude-haiku-4-5
Cheaper candidate preserved the policy, action, and explanation for ticket-summary case 20.