PRICING / STANDARD

Focused inference.
Simple pricing.

Price the decision layer by input tokens. No generated essay required.

DECOICE 1.0.0 · STANDARD

$0.042

per 1 million input tokens

Output-token price$0
Decision primitivesProbability · Choice · Score
AccessProvisioned API key
Integration guide ↗

Price and cost are
different numbers.

This is the chosen launch selling price. It is not a claim that our infrastructure costs less than every alternative at every utilization level. We may subsidize the difference.

The current launch is Standard mode served from the US. Response time depends on client region, input length and demand. Fast mode is not offered at launch.

Capacity is qualified, not unlimited.

API keys have explicit rate, concurrency and request-size limits. Requests beyond those limits can be rejected rather than queued without bound. The current early-access service can be interrupted. An uninterrupted-service guarantee is not offered.

The benchmark's estimated cost per request includes its stated resources and measurement window. Idle time, startup, redundancy and network charges must be accounted for separately when estimating production margins.