launch weekPro / Team is $0 / month $49 — full engine, no card charged.7 days left · ends Aug 20claim launch offer →
pricing

Pay for savings.
Never for the plumbing.

Start free with a developer key and run real traffic through the production engine. Paid tiers are priced against the compute dollars ACE reclaims — if it saves nothing, you owe nothing.

⚡ Zero-Downtime Guarantee — Never Breaks Production

When request limits are exceeded on any tier, skills automatically turn into shadow mode (quiet passthrough — zero actions taken on your requests). Eligible to re-activate in prod once limit refreshes or tier is upgraded.

Developer
$0/ month

A frictionless, top-of-funnel magnet — prove the value in your own terminal.

volume

10,000–50,000 requests / logs per month

  • Base-URL swap, all providers
  • Standard context compaction
  • Lightweight semantic caching
  • Dynamic intent routing
  • Live cost-savings proof of concept
estimated cost savings

Basic caching + routing immediately demonstrates ROI on repetitive queries — a live preview of the 30–80% savings unlocked by heavier pruning on paid tiers.

Pro / Team
$49/ month

The core self-serve revenue engine for established apps and tech leads.

volume

1,000,000 standard requests + 100k GPU-pruned requests

overage

* $0.50 per additional 1,000 pruned requests

$10.00 per additional 1,000,000 standard requests (or $0.01 per 1k)

  • Live telemetry dashboards
  • Full semantic caching + DB logging
  • Dynamic intent routing
  • Heavy context pruning (capped at 250k)
  • Transparent usage-based overage
estimated cost savings

Delivers the core promise: slash monthly API and GPU costs by 30–80% without sacrificing accuracy or speed — instantly reclaiming gross margins.

Enterprise
customfrom $1,000 / month

Massive scale and strict compliance — data never leaves your cloud.

volume

Massive enterprise volume

  • VPC / on-premise deployment
  • Advanced governance (RBAC, SSO, custom PII masking)
  • Custom SLAs
  • GPU fleet merit-order dispatch
  • Named engineer, shared Slack
estimated cost savings

Applies the 30–80% compute reduction at enterprise scale — shaving token bloat translates to massive absolute dollar savings while keeping all data secure.

Verified savings = provider list price minus net invoice, measured on your own traffic in the control plane.