Parleq for organizations
Back cloud cleanup with your corporate OIDC identity provider — Okta, Microsoft Entra ID, or any OIDC IdP — federated into AWS Bedrock or Google Vertex AI. Every employee dictates with the account they already have. No per-user API keys, no CLI dependence, no Parleq servers in the data path — ever.
Parleq itself costs nothing — there's no per-seat price and no license to buy. The only bill is your own cloud provider's charge for the LLM tokens, on the contract you already hold.
Trust model
One PKCE sign-in with your IdP backs cloud cleanup for everyone — not just the engineers with a CLI configured. The properties that matter before procurement:
Managed deployment
Push Parleq through your existing MDM — Jamf, Kandji, Mosyle, or Intune — with a managed configuration profile that pins the parts you care about and locks users out of changing them.
Pin the provider, restrict the model list, and pin the OIDC issuer and client ID plus the AWS role / GCP workforce provider. Users sign in — but they can't re-point the app at a personal tenant while the profile is in place.
No Parleq admin portal, no license-key system, no license-key check-in or usage beacon — the only call to a Parleq-controlled host is the Sparkle update check against parleq.app (version + OS string, no identifiers, no dictation data). The managed profile is the whole control surface.
For your security team
Parleq ships a security-review packet in the repo — data flows, trust boundaries, and audit history — so your security team can read it before procurement. And because Parleq is open source (Apache-2.0), reviewers can read the code itself rather than take a vendor's word for it.
docs/SECURITY_REVIEW.md on GitHub — data flows, trust boundaries, and the audit history. Procurement
No per-seat pricing, no license keys, no vendor contract to negotiate. Deploy to one Mac or ten thousand at the same price: zero.
The only cost is your own cloud provider's charge for LLM tokens, billed to the AWS or Google contract you already hold — not to Parleq.
Apache-2.0 and fully open source. Your security team verifies the claims against the source itself rather than trusting a vendor's word.
The setup playbooks: federate your IdP into AWS Bedrock and Google Vertex AI, leg by leg.
Setup playbooks →
The full schema for every managed-config key — provider pinning, OIDC pins, and the rest.
Schema reference →
Deploying to a fleet, step by step — the MDM profile, upload, and verify workflow.
Deployment workflow →
Free and Apache-2.0. Read the code, the security-review packet, and the audit history — and verify every claim on this page yourself.
Read the source →