Measured sequential-v2 run
Protocol dataset: 20 routing cases and six preselected multi-step planning cases, evaluated independently of routing. No real tools are executed. A plan is proposed one action at a time, with up to three actions followed by stop.
This report contains 20 unique cases and 156 case attempts across the selected tracks/providers/orders/repeats.
Models: jev: jev-1.13.0, openai: gpt-4.1-mini-2025-04-14
| Provider | Track | Completed / attempted | Strict route | Operational route | First action | Exact sequence | Median ms / case | API calls |
|---|---|---|---|---|---|---|---|---|
| jev | plan | 8/18 | N/A | N/A | 100.0% (n=18) | 33.3% (n=18) | 1195.334 | 68 |
| jev | route | 60/60 | 85.0% (n=60) | 90.0% (n=60) | N/A | N/A | 341.259 | 60 |
| openai | plan | 9/18 | N/A | N/A | 100.0% (n=18) | 16.7% (n=18) | 4533.702 | 71 |
| openai | route | 60/60 | 86.7% (n=60) | 86.7% (n=60) | N/A | N/A | 900.710 | 60 |
A planning case includes ALL its API calls, including the stop decision. Jev output is free at the saved rate. Missing usage remains unknown. Projected cost is per 1,000 cases/plans, not per 1,000 individual API calls.
| Provider | Track | Input cost | Output cost | Total cost | Projected / 1,000 cases | Cost coverage |
|---|---|---|---|---|---|---|
| jev | plan | 0.002188032 | 0.000000000 | 0.002188032 | 0.121557 | 18/18 |
| jev | route | 0.001633464 | 0.000000000 | 0.001633464 | 0.027224 | 60/60 |
| openai | plan | 0.014043600 | 0.000758400 | 0.014802000 | 0.822333 | 18/18 |
| openai | route | 0.009069600 | 0.000624000 | 0.009693600 | 0.161560 | 60/60 |
| Case | Provider | Track | Order | Repeat | Prediction / partial plan | Status | First correct | Sequence correct | Total ms | Calls |
|---|---|---|---|---|---|---|---|---|---|---|
| R11 | jev | plan | original | 1 | ['google_drive', 'gmail'] | ok | True | True | 1049.892 | 3 |
| R11 | openai | plan | original | 1 | ['google_drive', 'gmail', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 4264.708 | 4 |
| R11 | jev | plan | original | 2 | ['google_drive', 'gmail', 'google_drive', 'gmail'] | max_steps | True | False | 1352.263 | 4 |
| R11 | openai | plan | original | 2 | ['google_drive', 'gmail', 'general_reasoning'] | ok | True | False | 4602.432 | 4 |
| R11 | jev | plan | original | 3 | ['google_drive', 'gmail'] | ok | True | True | 1046.999 | 3 |
| R11 | openai | plan | original | 3 | ['google_drive', 'gmail', 'general_reasoning'] | ok | True | False | 4223.559 | 4 |
| R13 | jev | plan | original | 1 | ['crm', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1328.694 | 4 |
| R13 | openai | plan | original | 1 | ['crm', 'general_reasoning', 'general_reasoning'] | ok | True | False | 4666.030 | 4 |
| R13 | jev | plan | original | 2 | ['crm', 'general_reasoning', 'general_reasoning'] | ok | True | False | 1330.847 | 4 |
| R13 | openai | plan | original | 2 | ['crm', 'general_reasoning'] | ok | True | True | 3386.041 | 3 |
| R13 | jev | plan | original | 3 | ['crm', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1377.351 | 4 |
| R13 | openai | plan | original | 3 | ['crm', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 4040.864 | 4 |
| R17 | jev | plan | original | 1 | ['gmail', 'calendar'] | ok | True | True | 1059.822 | 3 |
| R17 | openai | plan | original | 1 | ['gmail', 'calendar', 'general_reasoning', 'gmail'] | max_steps | True | False | 5327.048 | 4 |
| R17 | jev | plan | original | 2 | ['gmail', 'calendar'] | ok | True | True | 1006.161 | 3 |
| R17 | openai | plan | original | 2 | ['gmail', 'calendar', 'general_reasoning', 'gmail'] | max_steps | True | False | 4184.756 | 4 |
| R17 | jev | plan | original | 3 | ['gmail', 'calendar', 'general_reasoning'] | ok | True | False | 1626.318 | 4 |
| R17 | openai | plan | original | 3 | ['gmail', 'calendar', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 5006.919 | 4 |
| R18 | jev | plan | original | 1 | ['google_drive', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1649.650 | 4 |
| R18 | openai | plan | original | 1 | ['google_drive', 'general_reasoning', 'general_reasoning'] | ok | True | False | 6333.648 | 4 |
| R18 | jev | plan | original | 2 | ['google_drive', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1395.632 | 4 |
| R18 | openai | plan | original | 2 | ['google_drive', 'general_reasoning', 'general_reasoning'] | ok | True | False | 5211.931 | 4 |
| R18 | jev | plan | original | 3 | ['google_drive', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1359.070 | 4 |
| R18 | openai | plan | original | 3 | ['google_drive', 'general_reasoning', 'general_reasoning'] | ok | True | False | 4476.057 | 4 |
| R19 | jev | plan | original | 1 | ['gmail', 'crm', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1383.994 | 4 |
| R19 | openai | plan | original | 1 | ['gmail', 'crm', 'general_reasoning'] | ok | True | True | 3880.281 | 4 |
| R19 | jev | plan | original | 2 | ['gmail', 'crm', 'general_reasoning'] | ok | True | True | 1476.138 | 4 |
| R19 | openai | plan | original | 2 | ['gmail', 'crm', 'general_reasoning'] | ok | True | True | 4533.702 | 4 |
| R19 | jev | plan | original | 3 | ['gmail', 'crm', 'general_reasoning'] | ok | True | True | 1379.107 | 4 |
| R19 | openai | plan | original | 3 | ['gmail', 'crm', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 4532.397 | 4 |
| R20 | jev | plan | original | 1 | ['database', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1485.194 | 4 |
| R20 | openai | plan | original | 1 | ['database', 'crm', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 4398.910 | 4 |
| R20 | jev | plan | original | 2 | ['database', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1311.963 | 4 |
| R20 | openai | plan | original | 2 | ['database', 'crm', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 4525.248 | 4 |
| R20 | jev | plan | original | 3 | ['database', 'general_reasoning', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 1298.281 | 4 |
| R20 | openai | plan | original | 3 | ['database', 'crm', 'general_reasoning', 'general_reasoning'] | max_steps | True | False | 4188.566 | 4 |
| R01 | jev | route | original | 1 | google_drive | ok | None | None | 335.687 | 1 |
| R01 | openai | route | original | 1 | database | ok | None | None | 790.536 | 1 |
| R01 | jev | route | original | 2 | google_drive | ok | None | None | 327.307 | 1 |
| R01 | openai | route | original | 2 | database | ok | None | None | 1033.887 | 1 |
| R01 | jev | route | original | 3 | google_drive | ok | None | None | 374.859 | 1 |
| R01 | openai | route | original | 3 | database | ok | None | None | 763.534 | 1 |
| R02 | jev | route | original | 1 | gmail | ok | None | None | 324.327 | 1 |
| R02 | openai | route | original | 1 | gmail | ok | None | None | 873.460 | 1 |
| R02 | jev | route | original | 2 | gmail | ok | None | None | 341.767 | 1 |
| R02 | openai | route | original | 2 | gmail | ok | None | None | 1219.036 | 1 |
| R02 | jev | route | original | 3 | gmail | ok | None | None | 326.542 | 1 |
| R02 | openai | route | original | 3 | gmail | ok | None | None | 989.000 | 1 |
| R03 | jev | route | original | 1 | calendar | ok | None | None | 354.702 | 1 |
| R03 | openai | route | original | 1 | calendar | ok | None | None | 902.813 | 1 |
| R03 | jev | route | original | 2 | calendar | ok | None | None | 311.833 | 1 |
| R03 | openai | route | original | 2 | calendar | ok | None | None | 773.961 | 1 |
| R03 | jev | route | original | 3 | calendar | ok | None | None | 402.037 | 1 |
| R03 | openai | route | original | 3 | calendar | ok | None | None | 1128.287 | 1 |
| R04 | jev | route | original | 1 | crm | ok | None | None | 377.571 | 1 |
| R04 | openai | route | original | 1 | crm | ok | None | None | 808.781 | 1 |
| R04 | jev | route | original | 2 | crm | ok | None | None | 317.234 | 1 |
| R04 | openai | route | original | 2 | crm | ok | None | None | 832.111 | 1 |
| R04 | jev | route | original | 3 | crm | ok | None | None | 334.593 | 1 |
| R04 | openai | route | original | 3 | crm | ok | None | None | 924.417 | 1 |
| R05 | jev | route | original | 1 | database | ok | None | None | 316.445 | 1 |
| R05 | openai | route | original | 1 | database | ok | None | None | 875.272 | 1 |
| R05 | jev | route | original | 2 | database | ok | None | None | 354.485 | 1 |
| R05 | openai | route | original | 2 | database | ok | None | None | 737.880 | 1 |
| R05 | jev | route | original | 3 | database | ok | None | None | 340.750 | 1 |
| R05 | openai | route | original | 3 | database | ok | None | None | 827.863 | 1 |
| R06 | jev | route | original | 1 | general_reasoning | ok | None | None | 392.646 | 1 |
| R06 | openai | route | original | 1 | general_reasoning | ok | None | None | 766.232 | 1 |
| R06 | jev | route | original | 2 | general_reasoning | ok | None | None | 338.154 | 1 |
| R06 | openai | route | original | 2 | general_reasoning | ok | None | None | 818.699 | 1 |
| R06 | jev | route | original | 3 | general_reasoning | ok | None | None | 351.372 | 1 |
| R06 | openai | route | original | 3 | gmail | ok | None | None | 819.672 | 1 |
| R07 | jev | route | original | 1 | gmail | ok | None | None | 348.952 | 1 |
| R07 | openai | route | original | 1 | gmail | ok | None | None | 986.977 | 1 |
| R07 | jev | route | original | 2 | gmail | ok | None | None | 366.468 | 1 |
| R07 | openai | route | original | 2 | gmail | ok | None | None | 1036.194 | 1 |
| R07 | jev | route | original | 3 | gmail | ok | None | None | 313.665 | 1 |
| R07 | openai | route | original | 3 | gmail | ok | None | None | 907.401 | 1 |
| R08 | jev | route | original | 1 | gmail | ok | None | None | 347.320 | 1 |
| R08 | openai | route | original | 1 | google_drive | ok | None | None | 958.790 | 1 |
| R08 | jev | route | original | 2 | gmail | ok | None | None | 332.688 | 1 |
| R08 | openai | route | original | 2 | google_drive | ok | None | None | 854.813 | 1 |
| R08 | jev | route | original | 3 | gmail | ok | None | None | 372.283 | 1 |
| R08 | openai | route | original | 3 | google_drive | ok | None | None | 930.343 | 1 |
| R09 | jev | route | original | 1 | crm | ok | None | None | 391.938 | 1 |
| R09 | openai | route | original | 1 | crm | ok | None | None | 1270.400 | 1 |
| R09 | jev | route | original | 2 | crm | ok | None | None | 356.449 | 1 |
| R09 | openai | route | original | 2 | crm | ok | None | None | 837.510 | 1 |
| R09 | jev | route | original | 3 | crm | ok | None | None | 405.568 | 1 |
| R09 | openai | route | original | 3 | crm | ok | None | None | 1591.959 | 1 |
| R10 | jev | route | original | 1 | database | ok | None | None | 369.357 | 1 |
| R10 | openai | route | original | 1 | database | ok | None | None | 962.232 | 1 |
| R10 | jev | route | original | 2 | database | ok | None | None | 309.894 | 1 |
| R10 | openai | route | original | 2 | database | ok | None | None | 1589.273 | 1 |
| R10 | jev | route | original | 3 | database | ok | None | None | 339.789 | 1 |
| R10 | openai | route | original | 3 | database | ok | None | None | 906.011 | 1 |
| R11 | jev | route | original | 1 | multi_tool | ok | None | None | 381.207 | 1 |
| R11 | openai | route | original | 1 | multi_tool | ok | None | None | 734.533 | 1 |
| R11 | jev | route | original | 2 | multi_tool | ok | None | None | 321.271 | 1 |
| R11 | openai | route | original | 2 | multi_tool | ok | None | None | 869.947 | 1 |
| R11 | jev | route | original | 3 | multi_tool | ok | None | None | 355.588 | 1 |
| R11 | openai | route | original | 3 | multi_tool | ok | None | None | 928.575 | 1 |
| R12 | jev | route | original | 1 | calendar | ok | None | None | 306.247 | 1 |
| R12 | openai | route | original | 1 | calendar | ok | None | None | 2159.998 | 1 |
| R12 | jev | route | original | 2 | calendar | ok | None | None | 360.457 | 1 |
| R12 | openai | route | original | 2 | calendar | ok | None | None | 767.145 | 1 |
| R12 | jev | route | original | 3 | calendar | ok | None | None | 337.316 | 1 |
| R12 | openai | route | original | 3 | calendar | ok | None | None | 930.107 | 1 |
| R13 | jev | route | original | 1 | crm | ok | None | None | 343.028 | 1 |
| R13 | openai | route | original | 1 | multi_tool | ok | None | None | 2295.700 | 1 |
| R13 | jev | route | original | 2 | crm | ok | None | None | 324.981 | 1 |
| R13 | openai | route | original | 2 | crm | ok | None | None | 738.193 | 1 |
| R13 | jev | route | original | 3 | crm | ok | None | None | 336.777 | 1 |
| R13 | openai | route | original | 3 | multi_tool | ok | None | None | 993.794 | 1 |
| R14 | jev | route | original | 1 | ambiguous | ok | None | None | 334.481 | 1 |
| R14 | openai | route | original | 1 | ambiguous | ok | None | None | 1116.336 | 1 |
| R14 | jev | route | original | 2 | ambiguous | ok | None | None | 337.654 | 1 |
| R14 | openai | route | original | 2 | ambiguous | ok | None | None | 965.162 | 1 |
| R14 | jev | route | original | 3 | ambiguous | ok | None | None | 481.307 | 1 |
| R14 | openai | route | original | 3 | ambiguous | ok | None | None | 2972.583 | 1 |
| R15 | jev | route | original | 1 | gmail | ok | None | None | 322.899 | 1 |
| R15 | openai | route | original | 1 | google_drive | ok | None | None | 994.447 | 1 |
| R15 | jev | route | original | 2 | multi_tool | ok | None | None | 334.304 | 1 |
| R15 | openai | route | original | 2 | google_drive | ok | None | None | 810.546 | 1 |
| R15 | jev | route | original | 3 | gmail | ok | None | None | 330.400 | 1 |
| R15 | openai | route | original | 3 | google_drive | ok | None | None | 879.362 | 1 |
| R16 | jev | route | original | 1 | database | ok | None | None | 320.786 | 1 |
| R16 | openai | route | original | 1 | database | ok | None | None | 865.889 | 1 |
| R16 | jev | route | original | 2 | database | ok | None | None | 330.050 | 1 |
| R16 | openai | route | original | 2 | database | ok | None | None | 1776.388 | 1 |
| R16 | jev | route | original | 3 | database | ok | None | None | 333.205 | 1 |
| R16 | openai | route | original | 3 | database | ok | None | None | 1121.951 | 1 |
| R17 | jev | route | original | 1 | multi_tool | ok | None | None | 584.409 | 1 |
| R17 | openai | route | original | 1 | multi_tool | ok | None | None | 815.346 | 1 |
| R17 | jev | route | original | 2 | multi_tool | ok | None | None | 341.862 | 1 |
| R17 | openai | route | original | 2 | multi_tool | ok | None | None | 879.830 | 1 |
| R17 | jev | route | original | 3 | multi_tool | ok | None | None | 354.443 | 1 |
| R17 | openai | route | original | 3 | multi_tool | ok | None | None | 811.718 | 1 |
| R18 | jev | route | original | 1 | multi_tool | ok | None | None | 330.681 | 1 |
| R18 | openai | route | original | 1 | multi_tool | ok | None | None | 898.608 | 1 |
| R18 | jev | route | original | 2 | multi_tool | ok | None | None | 352.854 | 1 |
| R18 | openai | route | original | 2 | multi_tool | ok | None | None | 834.214 | 1 |
| R18 | jev | route | original | 3 | multi_tool | ok | None | None | 341.944 | 1 |
| R18 | openai | route | original | 3 | multi_tool | ok | None | None | 854.301 | 1 |
| R19 | jev | route | original | 1 | multi_tool | ok | None | None | 479.424 | 1 |
| R19 | openai | route | original | 1 | multi_tool | ok | None | None | 963.048 | 1 |
| R19 | jev | route | original | 2 | multi_tool | ok | None | None | 324.998 | 1 |
| R19 | openai | route | original | 2 | multi_tool | ok | None | None | 855.415 | 1 |
| R19 | jev | route | original | 3 | multi_tool | ok | None | None | 343.826 | 1 |
| R19 | openai | route | original | 3 | multi_tool | ok | None | None | 1905.851 | 1 |
| R20 | jev | route | original | 1 | multi_tool | ok | None | None | 362.989 | 1 |
| R20 | openai | route | original | 1 | multi_tool | ok | None | None | 1040.589 | 1 |
| R20 | jev | route | original | 2 | multi_tool | ok | None | None | 330.279 | 1 |
| R20 | openai | route | original | 2 | multi_tool | ok | None | None | 874.562 | 1 |
| R20 | jev | route | original | 3 | multi_tool | ok | None | None | 354.715 | 1 |
| R20 | openai | route | original | 3 | multi_tool | ok | None | None | 833.600 | 1 |
Primary metrics use original order. Errors and step-limit failures count as incorrect, but a validated first action is scored independently. Median latency sums SDK-call times for each completed case (network included). Repeats are correlated; only six independent planning cases. See summary.json for consistency/order-stability coverage and results.jsonl for full prompts and raw responses. This revised harness is a new experiment, not directly comparable with the previous parallel-slot latency. No calibration or universal superiority claim is supported.