parelyxPricing
Plans for every stage. Priced with you.
Pricing is agreed with each customer. Plans describe scope and support — not request quotas.
Pricing on request
Developer
Engineers evaluating Parelyx on a single application.
Price
On request
- OpenAI-compatible integration
- Request-level observability
- Semantic reuse and prompt optimisation controls
- Community support
Growth
RecommendedTeams running production AI workloads across several applications.
Price
On request
- Everything in Developer
- Model and provider selection policies
- Workload baselining with our engineers
- Direct engineering support
Enterprise
Organisations with multiple teams, providers and governance requirements.
Price
On request
- Everything in Growth
- Deployment in your environment
- Security and procurement review
- Architecture review with our engineers
Final scope and terms are confirmed in your agreement.
Why are there no prices?
Every deployment differs in scope, environment and support needs, so we agree pricing directly with each customer rather than publishing a one-size rate card.
What determines which plan fits?
Mainly the number of applications and teams involved, your deployment requirements, and the level of engineering support you need — not a request quota.
Will Parelyx reduce our inference costs?
It depends on your traffic: how repetitive it is, how much context it carries and how it is currently routed. We start by baselining your workloads so that any change is measured, not promised.
Where does Parelyx run?
In your environment, alongside your applications and models. Deployment topology, data retention and access controls are agreed during onboarding.