Everything between a person and a model.

Identity and access Single sign-on with OIDC Sign in with Okta, Entra ID, Authentik, Keycloak or AD FS, with accounts and groups kept in step with your directory. How it works → Identity and access SCIM 2.0 provisioning Business Your identity provider pushes people and groups to Janus, and leavers lose access, keys and sessions at once. How it works → Identity and access Explicit model grants Signing in grants nothing. You decide which person, group, team or service may use each model. How it works → Janus Chat Janus Chat A chat app in the browser for every model a person is granted, where each message is checked, limited and metered like any API call. How it works → Models, providers and routing One OpenAI-compatible API Any OpenAI SDK or tool works with Janus by changing the base URL and the key. How it works → Models, providers and routing Hosted and self-hosted providers Cloud APIs and your own GPU servers behind one endpoint: OpenAI, Anthropic, Bedrock, Vertex, Gemini, vLLM, Ollama and more. How it works → Models, providers and routing Managed model aliases Multiple editions Publish a stable name like chat-default, then repoint it to a better or cheaper model without touching a single app. FreeAliases: publish, grant and repoint+ BusinessAutomatic fallback and load-balanced pools How it works → Models, providers and routing Load-balanced model pools Business Spread one model across several GPU servers by live queue length, and keep each conversation on the server that holds its cache. How it works → Models, providers and routing Auto-routing Callers ask for one router name, and a classifier sends each request to the model you chose for that kind of work. How it works → Models, providers and routing Decision models Classifier models that answer typed questions with probabilities, used inside Janus and served to people and apps. How it works → Models, providers and routing Personal subscriptions People bring the ChatGPT, Copilot, Grok or Mistral plan they already pay for, and it becomes metered and policy-checked. How it works → Metering, cost and limits Metering that matches the invoice Every request is metered by input, output and cache tokens at the provider's real prices, with no prompts stored. How it works → Metering, cost and limits Quotas and budgets Cap tokens, spend or requests per person, team or service, with warnings before anyone hits the limit. How it works → Dashboards and reports Reporting workspace Multiple editions Eleven report templates and a builder, with exports to PDF, Excel, CSV and JSON. FreeTemplates, builder, saved reports and every export format+ BusinessScheduled reports How it works → Dashboards and reports Usage insights See what kinds of work people use AI for, from category labels on sampled requests, without storing prompts. How it works → Security gateway Security policies and bindings Multiple editions Bundle checks into policies and bind them to everyone, a group, a model, a provider or a service token. FreePolicies and bindings in observe mode+ BusinessRedact and block modes How it works → Security gateway Secret detection and redaction Multiple editions Twenty-seven detectors catch API keys, cloud credentials and passwords, then redact them or block the request. FreeDetect and report all 27 secret types+ BusinessRedact the secret or block the request How it works → Security gateway Child Safety Rules that follow each child to every model they can use: age ranges, a tutor that never gives the answer, kind replies when they feel low, and a note to you when it matters. How it works → Security gateway Streamed answers checked in flight Multiple editions Streamed answers are checked a few words ahead, so a secret can be removed before it reaches the screen. FreeScan streamed answers and report+ BusinessRedact in flight or cut the stream How it works → Deployment and operations High availability Business Run several stateless replicas against one PostgreSQL database, so the gateway is never a single point of failure. How it works → Deployment and operations Air-gap ready Multiple editions Every edition installs and verifies its license with no network access, for sites that can't call home. FreeOffline install, offline license check, JANUS_OFFLINE+ EnterprisePerpetual keys, purchase by PO, air-gap support under an SLA How it works →