Changelog
What's built, and what's next.
Every feature is built and tested on parezen.com. Request an invite and start with the savings forecast.
Pricing updated 4 Nov 2026
Where things stand
- Built
The app, prices and fee rules
Every dashboard and admin page, the published prices and the savings-share rules. The sample organisation lets you click through with sample numbers.
- Built
Backend, gateway and integrations
The backend, the hosted gateway, Excel and CSV upload and the savings forecast, read-only billing connectors, the ledger and bill match, invoices through Paddle, SSO and SCIM. Provider keys are accepted only while encryption and rate limiting are on.
Latest updates
Self-hosted
Run the gateway in your own cloud on Docker or Kubernetes: a $30K a year licence plus 15% of savings. Prompts and provider keys never leave your network.
Pricing on your own keys
Free to $10K a month at list price, then 3%, never more than we save you on the traffic we optimise, or the savings share instead.
Sign-up in three steps
Three steps from a new organisation to your first savings estimate, with no code change. Sign-up stays invite-only until launch: request an invite on the sign-up page.
Everything launches together
Parezen is invite-only while we finish launch. Then every module, connector and the gateway open at once on parezen.com, with billing through Paddle. Self-hosted options (Docker, Azure, AWS, Google Cloud, Kubernetes) follow later.
Excel upload and Paddle payments
The savings forecast now reads Excel workbooks as well as CSV and JSONL. Invoices are paid by card through Paddle, our reseller, or by bank transfer. Cache and batch savings count on the ledger from the day you turn them on.
The backend is built and tested
The dashboard's backend and the hosted gateway are built and covered by tests: routing, caching, budgets and hard stops, shadow tests, the savings ledger and bill match, invoices, alerts, single sign-on (OpenID Connect and SAML) with SCIM, and per-project content logging.
A way out if Parezen is unreachable
The fail-open guide shows how your app can send a call straight to your provider when Parezen cannot be reached or does not answer in time. Budget hard stops are never bypassed.
Hosted first
Parezen runs the gateway at api.parezen.com. You change one URL and bring your own provider keys, which Parezen stores encrypted.
Launching with the full feature set
There is no pilot programme. Parezen opens once every part of the loop is ready. Request an invite to start with the savings forecast.
Published prices
Free under $1,000 a month of AI spend, then FinOps, Inference and Governance by tracked spend, with an optional share of verified savings.
What launches
Every feature on the homepage, grouped by the step it belongs to. All of them are built.
See
- Connect read-onlyBuilt
- Savings forecastBuilt
- Savings recommendationsBuilt
- Right-size found in the savings forecastBuilt
- Prompt caching found in the savings forecastBuilt
Cut
- Gateway applies a fixBuilt
- Automatic model routingBuilt
- Prompt caching applied by the gatewayBuilt
- Output controlBuilt
- Batch async workBuilt
- Cross-provider switchingBuilt
Prove
- Savings ledgerBuilt
- Baseline countBuilt
Run
- Hosted gateway (api.parezen.com)Built
From your first audit to how you buy AI compute.
The first five are built and launch together. Buying AI compute comes later.
Savings forecast
See where the money goes, from a usage export or a read-only connector.
Verified savings
Every saving counted against your baseline, then matched to the provider bill.
Cost per task
Agent and task budgets, priced per outcome instead of per token.
FinOps and governance
One place for spend, budgets, keys and policy across every provider.
Self-tuning routing
Routing that improves itself, inside the rules and quality bar you set.
Buying AI compute
Pooling demand so teams pay less for the same capacity. A later decision.