Quote and choose parcel carriers for a batch of shipments under each carrier's own rules
Exact prompt
# Quote and choose parcel carriers for a batch of shipments under each carrier's own rules Inputs: `carriers.json` (each carrier's dimensional divisor, weight rounding, billing method, zones, rate bands and surcharges) and `shipments.csv` (shipment_id, piece, dest_postcode, dims_cm as "a x b x c" in any order, actual_kg). Everything is synthetic. `carriers.json` and these rules are the record; where `dispatch-note.txt` disagrees, the rules win. Rules: - A piece's chargeable weight with a carrier is the larger of actual_kg and a × b × c / dim_divisor, rounded up to the next multiple of the carrier's round_kg. - The zone is the one for the longest prefix in zones that the postcode starts with (compare in capitals without spaces). The base comes from the first band whose max_kg is at least the weight; above the last band it is the last band's price plus per_kg_above × the kg over its max_kg. billing per_piece prices each piece's chargeable weight separately and adds them; billing consignment prices the sum of the pieces' chargeable weights once. - Oversize applies per piece when its longest side is more than longest_cm, or its longest side plus twice the sum of the other two sides is more than length_plus_girth_cm. The remote fee applies once per shipment when the postcode starts with any remote prefix. - Fuel is fuel_pct of the sum of the charge types listed in fuel_applies_to (base, oversize, remote), rounded half-up to the cent. The total is base + oversize fees + remote fee + fuel. - A carrier can't take a shipment if any piece's actual_kg is above its max_piece_kg; its quote is then null. - Choose the cheapest carrier for each shipment; on a tie take the carrier listed first in carriers.json (null if no carrier can take it). - Write `out/quotes.json`: `{"chargeable_kg": {"<shipment_id>": {"<carrier>": [<kg per piece in piece order>], ...}, ...}, "base": {"<shipment_id>": {"<carrier>": <base or null>, ...}, ...}, "quotes": {"<shipment_id>": {"<carrier>": <total or null>, ...}, ...}, "chosen": {"<shipment_id>": "<carrier>", ...}}`. ## How this run works - Work on your own: nobody will answer questions in this task, and writing to `ask/` counts against the run. - Write only under `out/`. Create `out/` if it is missing. Inputs are synthetic; don't fetch anything from the network. - Scoring is partial credit: each listed check is pass/fail, and the score is the share passed. Numbers are compared to within 0.005 unless the task says otherwise.
Price a batch of synthetic multi-piece shipments with three carriers and choose the cheapest. Each carrier has its own dimensional divisor, round-up step and maximum piece weight; zones come from the longest postcode prefix; one carrier bills the whole consignment while the others bill per piece; bands run out and switch to a per-kg rate; oversize is judged per piece by length and girth; the remote fee applies once; and fuel is charged on different parts of the bill for each carrier. A dispatch note gets the billing, rounding, fuel and choice rules wrong. Original benchmark task written for usecaserank.com (not based on a specific public post). Headless, offline run on a reproducible fixture. Each run's notes say how its cost was measured (estimates are marked).
Success criteria: chargeable_weight zones_bands_billing surcharges_and_fuel eligibility choice
Best model for this task so far: Claude Haiku 5.5 (effort default (no --effort flag) · thinking on) (success, $0.0088, 45s).
All runs (3)
100% pass rate across 3 eligible runs: 3 pass, 0 partial, 0 fail. Partial and failed runs are listed below with everything else.ⓘ
| Model | Cost | Time | Step-insⓘ | Autonomy | Outcome | Trustⓘ | Date | Compare |
|---|---|---|---|---|---|---|---|---|
| AClaude Opus 5.5default · thinking on | $0.18 | 32s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-24 | |
| AClaude Haiku 5.5default · thinking on | $0.0088 | 45s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-24 | |
| AClaude Sonnet 5.5default · thinking on | $0.07 | 15s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-24 |
Embed this result
A live badge for this use case. It links back here and updates as new runs are published.
Get new runs and ranking changes weekly
Which models got better or cheaper on real tasks like this one.