Model economics / detail
claude-4.8-opusD
Dispatch tier: Expensive (complexity class this model is routed for)
Three misrouted dispatches under a name that doesn't exist, one lucky success, 25s median. This isn't a model record, it's a typo with a scoreboard entry. Fix the router; judge the real claude-opus-4-8 instead.
The measured record
What the ledger shows
Fleet-ledger aggregates for this model only — run counts, outcomes, spend, and token appetite. Missing telemetry reads “Not observed”, never zero.
Agent runs
3
Success rate
33%
Metered spend
Not observed
cost-blind: no run was metered
Cost / metered run
Not observed
Tokens in
Not observed
Tokens out
Not observed
Tokens in / run
Not observed
Tokens out / run
Not observed
Outcomes
Status breakdown
Every run ends in exactly one status. The bar is the whole record, to scale.
- failure: 2 of 3 runs (66.7% of total)
Run duration
How long the runs take
Wall-clock distribution across measured runs, from the fastest exit to the longest grind.
Min
19s
p25
22s
Median
25s
p75
39s
p90
48s
p99
53s
Max
53s
Mean
32s
Duration measured on 3 of 3 runs.
Commentary
Idiosyncrasies
What stands out in this model's numbers — shape, appetite, and failure habits.
- Three runs under a name that is claude-opus-4-8 with its digits shuffled — a router typo with a scoreboard entry.
- 25-second median and one lucky success out of three; these runs never really started.
Commentary
Lessons learned
Practical routing and operations takeaways, grounded in the same record.
- Nothing to learn yet beyond: fix the router.
- Its "expensive" tier label is inherited from the name it was mistyped from, not from anything it did.