The weekly digest
A weekly digest your team actually forwards.
This is a sample with representative numbers. The real one lands in your inbox every Monday, built from your own routed traffic.
a sample team saved, last 30 days
$2,680
saved 68.4%
Actual spend through IQ Routing versus pinning every request to a frontier model (Opus / GPT-5 class).
You paid
$1,240
Frontier-pinned
$3,920
Requests
184,000
Cache hit rate
41.0%
billed at $0
Where the spend went
| Model | Cost | Requests | Share |
|---|---|---|---|
| gpt-4o-mini | $388.00 | 121,000 | 31.3% |
| claude-haiku-4-5 | $332.00 | 41,000 | 26.8% |
| gpt-4o | $304.00 | 16,500 | 24.5% |
| claude-opus-4-8 | $216.00 | 5,500 | 17.4% |
The frontier-pinned baseline prices the same input and output tokens at a frontier model (Opus / GPT-5 class)'s published rates with no caching. The actual figure is metered routed plus cached spend. The baseline is an estimate for comparison, not a quote.
IQ Routing
last 30 days, generated 2026-07-06