spoof
One key for every model. Route Claude, GPT and more through a single endpoint, with controls on every token.
Everything you need to route models.
A single base URL fronts every provider, with billing, failover and controls handled for you.
One endpoint, every model
Route Claude, GPT and more through a single base URL. Swap models without touching your integration.
Pay per call
You pay for the requests you make, nothing else. No subscriptions, no seats, no minimums.
Automatic failover
A provider stumbles, the request reroutes instantly. Never a dropped call.
Full metering
Every token tracked and attributed, down to the individual request.
Models on tap
Claude, GPT and more, ready to call. New models land as they ship.
Pay per call, every model
Final price per million tokens, settled in USDC on Base. Pay only for what you use, with no subscriptions or minimums.
claude-fable-5Input
$3.50 /M
Output
$17.50 /M
Cached
$0.35 /M
claude-haiku-4-5Input
$0.35 /M
Output
$1.75 /M
Cached
$0.04 /M
claude-opus-4-6Input
$1.75 /M
Output
$8.75 /M
Cached
$0.18 /M
claude-opus-4-7Input
$1.75 /M
Output
$8.75 /M
Cached
$0.18 /M
claude-opus-4-8Input
$1.75 /M
Output
$8.75 /M
Cached
$0.18 /M
claude-sonnet-4-6Input
$1.05 /M
Output
$5.25 /M
Cached
$0.11 /M
claude-sonnet-5Input
$1.05 /M
Output
$5.25 /M
Cached
$0.11 /M
Prices are per 1,000,000 tokens. Prompt-cache writes are billed at 1.25x the input rate (5-minute cache) or 2x (1-hour cache).
Questions, answered
Route your first request today.
One key, every model, billed by the request. No subscriptions, no setup.