Stable AI API plans

Subscription for daily availability. Compute Credits for advanced bursts.

TrexAPI pricing is designed around reliability: daily quota covers Smart Auto, and prepaid balance covers Premium models, pinned providers, long context, and high concurrency.

Smart Autodefault stable routing included in daily quota

Creditsprepaid balance for Premium and bursts

Fallbackmulti-provider resilience

You are buying an AI availability layer

Pricing is organized around stable API access, automatic fallback, quota recovery, balance control, and risk boundaries.

Subscriptions include daily quota

Smart Auto is the default stable path for ordinary traffic. Daily quota resets keep normal usage from being exhausted by one burst.

Compute Credits cover advanced usage

Premium models, pinned providers, very long context, excess concurrency, and bursts consume prepaid balance.

Availability-first downgrade behavior

When balance is insufficient, ordinary requests can downgrade to Smart Auto; Premium-required calls return `insufficient_balance`.

Plan overview

Self-serve stays simple: validate for free, then move to Stable Starter, Stable Pro, or Stable Scale. Enterprise covers contracts, security review, and private boundaries.

Developer

For validating the Stable AI API path before production traffic.

$0/ month

1 production API key
Smart Auto trial traffic
Basic dashboard
Cold-start quota and abuse controls

Start free

Stable Starter

Early access

For early users who want a stable fallback interface at a low monthly entry point.

$9.9/ month

Daily Smart Auto quota
API key management
Provider binding
Basic logs
Early supporter pricing while active

Choose Starter

Stable Pro

The default production plan for teams that need one endpoint, automatic routing, and reliable fallback.

$40/ month

Higher daily Smart Auto quota
Multi-provider fallback
Compute Credits top-up support
Priority email support
Usage and quota visibility

Choose Pro

Stable Scale

For heavier workloads that need larger quota, higher concurrency, and tighter support.

$120/ month

Large daily Smart Auto quota
Higher RPM and concurrency envelope
Premium and pinned-provider usage through credits
Priority support
Rollout planning

Choose Scale

Enterprise

For organizations that need contracted capacity, security review, private routing rules, or SLA terms.

Custom contract

Custom quota and rate envelope
Security and abuse-review workflow
Optional private deployment or SLA
Dedicated onboarding
Contract billing

Contact sales

Feature comparison

A quick view of free validation, self-serve production, and enterprise contracts.

Capability	Developer	Stable Starter	Stable Pro	Stable Scale	Enterprise
Primary use	Evaluation	Small production fallback	Default production API	High-volume production	Contracted platform
Smart Auto quota	Cold-start quota	Daily included quota	Higher daily quota	Large daily quota	Custom
Compute Credits	Top-up when available	Prepaid top-up	Prepaid top-up	Prepaid top-up + higher bursts	Contracted or prepaid
Fallback	Basic validation	Included	Included	Included + higher concurrency	Custom routing policy
API keys	1 key	Dashboard managed	Dashboard managed	Dashboard managed	Custom policy
Controls	Cold-start limits	RPM + quota controls	RPM + quota + balance controls	Higher risk envelope	Custom risk review
Support	Self-serve docs	Email	Priority email	Priority support	Sales + onboarding

Compute Credits usage rules

Smart Auto consumes daily quota by default.
Premium, pinned models, pinned providers, long context, and high concurrency consume credits.
If balance is insufficient, normal requests downgrade to Smart Auto; forced Premium requests return `insufficient_balance`.
Premium access opens after recharge to reduce payment and abuse risk.

Fair use and risk controls

All plans are subject to API key limits, RPM, concurrency, daily quota, IP/ASN risk checks, failed-payment handling, and sensitive-abuse detection. Higher capability opens progressively instead of being exposed without limits.

Stability first, then scale.

Connect Stable AI API first, confirm one stable production path, then enable Premium and higher concurrency as needed.