What it costs.

A daily view of HiveXS spend across jobs. Updated end-of-day from the Anthropic Admin API and our local activity log.

Last updated: 2026-05-09 12:27 AEST

HiveXS runs mostly on cloud AI models. Every call costs something. We track each dollar and attribute it back to the task that triggered it. Three numbers worth watching: total spend this month, spend per task, and spend per client.

Figures shown in USD primary with AUD conversion alongside. Rate 1.3805 (fetched 2026-05-09)
$603.83 ~AUD $833.58
Spend this month
Trajectory: $2,046.33 ~AUD $2,824.92 / month at current daily average
$555.87 ~AUD $767.37
Last 7 days
Daily average: $68.21 ~AUD $94.16
$613.90 ~AUD $847.48
Total tracked
Across 9 days of activity

Daily spend

Day by day, total cost and how many distinct tasks the spend is split across.

DateSpendTasks active
2026-05-08$91.45 ~AUD $126.255
2026-05-07$194.43 ~AUD $268.413
2026-05-06$102.51 ~AUD $141.517
2026-05-05$88.54 ~AUD $122.235
2026-05-04$24.05 ~AUD $33.201
2026-05-03$42.29 ~AUD $58.3810
2026-05-02$12.60 ~AUD $17.396
2026-05-01$47.96 ~AUD $66.218
2026-04-30$10.06 ~AUD $13.891

Spend by client

Where the money is going. HiveXS-comms = website + content + decks. HiveXS-system = infrastructure + runtime. _orphan = activity hours where no task LOG event was written; close that gap by extending LOG coverage.

ClientTotalTasks
HiveXS-comms$261.01 ~AUD $360.3210
_orphan$140.60 ~AUD $194.101
HiveXS-system$134.93 ~AUD $186.2715
SHWR$43.73 ~AUD $60.377
AUSM$33.62 ~AUD $46.413

Top 15 tasks by spend

Tasks ranked by cost. Useful for spotting the heaviest jobs and asking whether they are worth what they are costing.

TaskClientCostInput tokensOutput tokens
T-2026-05-06-006-hivexs-website-v1HiveXS-comms$205.56 ~AUD $283.7748.5k1120.1k
T-2026-05-06-003-hivexs-knowledge-architecture-phase-2-buildHiveXS-system$49.37 ~AUD $68.151.1k248.6k
T-2026-05-05-003-showerloop-favicon-fixSHWR$22.96 ~AUD $31.700.7k76.3k
T-2026-05-08-001-hivexs-cost-accounting-designHiveXS-system$19.89 ~AUD $27.462.4k125.3k
T-2026-05-06-001-hivexs-knowledge-architectureHiveXS-system$18.50 ~AUD $25.540.8k117.2k
T-2026-05-05-002-ausmade-strategy-doc-proofAUSM$17.11 ~AUD $23.620.7k83.0k
T-2026-05-07-001-linkedin-reply-colin-stevensHiveXS-comms$12.69 ~AUD $17.521.9k45.6k
T-2026-05-04-004-ausm-onboardingAUSM$10.77 ~AUD $14.870.4k123.9k
T-2026-05-01-001-focus-group-cause-effect-v2HiveXS-comms$10.24 ~AUD $14.140.2k197.9k
T-2026-05-08-002-hivexs-local-model-strategyHiveXS-system$10.24 ~AUD $14.140.7k66.9k
T-2026-05-05-001-email-phase-3-inboundHiveXS-system$9.30 ~AUD $12.840.3k106.5k
T-2026-05-01-006-showerloop-onboardingSHWR$8.91 ~AUD $12.300.2k102.1k
T-2026-05-03-009-system-team-summary-deck-v2HiveXS-comms$7.57 ~AUD $10.450.1k49.8k
T-2026-05-03-008-system-team-summary-deckHiveXS-comms$7.41 ~AUD $10.230.2k41.0k
T-2026-05-06-004-ausm-edm-clean-html-exportAUSM$5.75 ~AUD $7.940.2k58.2k

Tokens by model

Where the tokens flowed. Cache reads dominate cost on opus-class models because long-running conversations re-read the cached context every turn.

ModelInputOutputCache readCost
claude-opus-4-774.3k4405.7k1001548.5k$611.29 ~AUD $843.88
claude-sonnet-4-60.1k100.1k3696.4k$2.61 ~AUD $3.60

How attribution works

A practical note on the numbers. Anthropic's Admin API gives us hourly token usage per model. We split that hour's cost across whichever HiveXS tasks were active in the same hour, weighted by LOG event count. Tasks with no LOG activity get nothing; orphan hours (no task active) go to the _orphan bucket. Approximate but directionally accurate. Phase 2 (per-call attribution at the OpenClaw runtime layer) would make this exact.