Skip to content

Claude Sonnet 5.5 (effort default (no --effort flag) · thinking on) on Charge daily late-payment interest and fixed compensation on overdue business invoices with part payments and credit notes

AClaude Sonnet 5.5default · thinking on · Claude Code 2.1.299 (headless -p) · cloud · 2026-10-22 · by @bobsiopener

Trust 80%Fully autonomoussuccess · first try
Estimated cost
$0.06
Wall time
15s
Human step-ins
0
Outcome
Pass

Compare with other models on this task →

Summary

Exact prompt

# Charge statutory interest and fixed compensation on late invoices Inputs: `invoices.csv` (every invoice issued to business customers, with issue and due dates), `payments.csv`, `credit-notes.csv` and `terms.json` (the annual interest rate, the statement date and the fixed-compensation bands). Work out the late-payment charges as at the statement date: - Interest runs on each day from the day after the due date up to and including the statement date, or up to and including the day the invoice was paid in full if that is earlier. - Each day's interest = outstanding principal x annual rate / 365. A payment reduces the principal from the day after it is received. - A credit note reduces the invoice amount from the start, as if that part had never been billed, whatever its date. - Fixed compensation: one fee per invoice that was paid in full after its due date or is still unpaid after it, chosen by the invoice amount after credit notes from the `fixed_fees` bands (the first band whose `below` is greater than that amount; `null` means no upper limit). - Round each invoice's interest to the cent half-up; totals are sums of the rounded figures. Write `out/late-charges.json`: `{"interest": {"<invoice_id>": <interest>, ...}, "fees": {"<invoice_id>": <fee>, ...}, "total_interest": <...>, "total_fees": <...>}`. List only invoices that have at least one day of interest, in both maps. ## 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.

Work out statutory late-payment interest and a banded fixed fee for every overdue invoice as at a statement date. Interest runs daily from the day after the due date, part payments cut the principal from the next day, and a credit note counts from the start even when it was issued after the due date. Starting a day early, ignoring part payments or applying a late credit note from its own date each change the charges. 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: interest_days part_payments credit_notes fixed_fees

Setup notes: Batch 23 case y02-late-payment-interest (offline, synthetic inputs, seed 681916460). Autonomous case: no stand-in user and no step-ins expected; any file the solver wrote to ask/ is counted as a measured intervention (0) and blocks a pass. Prompt = TASK.md verbatim. Hidden key kept outside the solver's work dir. Scored with the case's scorer (python3 score.py WORKDIR KEYDIR, absolute paths); pass = every check true and no ask/ files. Tools: Read/Write/Edit/Glob/Grep/Bash, web disabled. Single run, no retries. 1-min load average at scoring 0.20 (8 cores). Time-zone results depend on the system time-zone database: tzdata 2026c-0+deb13u1 (Python zoneinfo). Claude Code's own reported cost for claude-sonnet-5.5: $0.059398. Score: 4/4. Cost is an ESTIMATE: measured tokens x pricing.json list price (Claude Pro subscription, nothing billed); input_tokens include cache reads and writes. Single run.

Estimated cost List-price estimate; not actually billed.

Metrics

Estimated cost
$0.06
Wall time
15s
Human step-ins
0
Turns
4
Model usage
ModelInput tokOutput tokCost
AClaude Sonnet 5.5anthropic/claude-sonnet-5.5effort default (no --effort flag) · thinking on79,2051,667$0.06

Reproduce this run

  1. Use the use case prompt on the use case page.
  2. Same setup: Claude Sonnet 5.5 (anthropic/claude-sonnet-5.5, effort default (no --effort flag) · thinking on) via Claude Code 2.1.299 (headless -p), cloud.
  3. It counts as a success only if it meets every success criterion: interest_days part_payments credit_notes fixed_fees
  4. Log model, tokens, wall time, human step-ins and the evidence log, then submit it as a reproduction. A different model? Submit a head-to-head rerun.

How runs are scored: Methodology.

Was this useful?

⇄ Rerun on another model

Automated checks

Trust 80% comes from these checks (how Trust % works).

Evidence log
[
 {
  "duration_api_ms": 13537,
  "stop_reason": "end_turn",
  "session_id": "[REDACTED]",
  "total_cost_usd": 0.05939760000000001,
  "usage": {
   "input_tokens": 8,
   "cache_creation_input_tokens": 8921,
   "cache_read_input_tokens": 70276,
   "output_tokens": 1667,
   "output_tokens_details": {
    "thinking_tokens": 233
   },
   "server_tool_use": {
    "web_search_requests": 0,
    "web_fetch_requests": 0
   },
   "service_tier": "standard",
   "cache_creation": {
    "ephemeral_1h_input_tokens": 8921,
    "ephemeral_5m_input_tokens": 0
   },
   "inference_geo": "global",
   "iterations": [
    {
     "input_tokens": 2,
     "output_tokens": 465,
     "cache_read_input_tokens": 20124,
     "cache_creation_input_tokens": 1336,
     "cache_creation": {
      "ephemeral_5m_input_tokens": 0,
      "ephemeral_1h_input_tokens": 1336
     },
     "type": "message"
    }
   ],
   "speed": "standard",
   "fallback_credit": null
  },
  "modelUsage": {
   "claude-sonnet-5-5": {
    "inputTokens": 8,
    "outputTokens": 1667,
    "cacheReadInputTokens": 70276,
    "cacheCreationInputTokens": 8921,
    "webSearchRequests": 0,
    "costUSD": 0.05939760000000001,
    "contextWindow": 1000000,
    "maxOutputTokens": 128000,
    "thinkingTokens": 233,
    "canonicalModel": "claude-sonnet-5-5",
    "provider": "firstParty",
    "costBasis": "list"
   }
  },
  "permission_denials": [],
  "terminal_reason": "completed",
  "fast_mode_state": "off",
  "fast_mode_disabled_reason": "sdk_opt_in_required",
  "subagent_stats": {
   "spawned": 0,
   "requested": {
    "background": 0,
    "foreground": 0,
    "unset": 0
   },
   "started_in_background": 0,
   "max_depth": 0,
   "spawned_by_subagents": 0,
   "completed": 0,
   "failed": 0,
   "killed": {
    "parent": 0,
    "user": 0,
    "system": 0
   },
   "refused": {
    "depth_limit": 0,
    "concurrency_limit": 0,
    "budget": 0
   },
   "by_type": {}
  },
  "safety_stops": 0,
  "is_error": false,
  "num_turns": 4,
  "subtype": "success",
  "api_error_status": null,
  "result": "I wrote `out/late-charges.json`. 12 invoices have at least one day of interest. Total interest is 681.97 and total fixed fees are 750.00. I didn't check any figure by hand.\n\nI ran the calculation as a script over the four inputs. I read the 12.25 rate as a percentage, so the daily rate is 0.1225/365.\n\n- **Interest:** Each day's interest is the outstanding principal \u00d7 rate / 365. That principal is the invoice amount less credit notes, minus any payments dated before that day. The last interest day is the statement date, or the day the invoice was paid in full if that is earlier.\n- **Fees:** Every invoice with interest also gets a fee, because each one was either paid in full after its due date or is still unpaid. The fee comes from the `fixed_fees` band matching the amount after credit notes.\n- **Rounding:** Each invoice's interest is rounded half-up to the cent, and the totals are sums of those rounded figures.\n\nOne example: INV-4609 has a credit note of 4,208.72, which leaves 959.94 to pay. Two payments of 479.97 cleared it on 2026-09-26, so interest stops on that day. That gives 9.99 interest and the 40.00 fee.",
  "ttft_ms": 2552,
  "type": "result",
  "duration_ms": 13904,
  "uuid": "[REDACTED]",
  "ttft_stream_ms": 1727,
  "time_to_request_ms": 255,
  "first_content_frame_ms": 1727,
  "queued_turn_count": 0,
  "result_index": 0
 }
]

Get new runs and ranking changes weekly