Skip to main content
This page describes the onboarding journey for enterprise customers: the milestones between signature and production, what your team owns at each one, and how OpenRouter supports you. A typical enterprise onboarding tracks toward three milestones:
  • API keys, within 24 hours of signature
  • First inference from your own environment, within 48 hours of signature
  • Go-live by day 14, and full production pace by day 30
These dates describe the typical path rather than a guarantee. The dates that apply to you are agreed in your implementation plan. Progress through these milestones depends on a small number of inputs from your side, listed under each phase. The remaining steps are coordinated side by side with your team by the OpenRouter roles assigned to your account. Each account has two customer-facing roles assigned to it. The Forward Deployed Engineer (FDE) owns Activation and Onboarding through go-live. The Customer Success Manager (CSM) joins from Onboarding and owns the relationship after go-live.
OpenRouter provisions your account, credits, and invoicing from its side, but cannot access or configure anything inside your organization: workspaces, API keys, guardrails, members, or settings. Your team retains access and performs all configuration. During working sessions, your FDE works side by side with your team to get everything set up.
If you want to set up your organization self-serve, follow the Enterprise Quickstart. After go-live, the cadence is the monthly and quarterly reviews described under Optimization.

Milestones at a glance

The two live sessions are the day-one activation call and the day-five-to-seven inference and observability session, while everything else in this window runs asynchronously in the shared channel.

Before you sign

Two working sessions happen before signature so that no time is spent on them afterwards. Both are led by OpenRouter and take about 30 minutes each. Security and governance session. Data retention and Zero Data Retention enforcement, provider logging, in-region routing, SSO and SCIM requirements, DPA, SOC 2, and guardrail policy. You receive written answers to every question raised, so your security reviewers have a document rather than meeting notes. Technical scoping session. Your use case, expected month-one volume, integration surface, outbound network restrictions, and the observability platform you want traces delivered to. You then receive a dated implementation plan, with your named owners against each milestone, before you sign. Your activation call is scheduled before signature and booked for the first business day after signature.

What we need from you

Activation: Day 0 to 2

Milestone: your first production inference, within 48 hours of signature. Your team creates and holds its own API keys throughout. OpenRouter never creates keys on your behalf.

Day 0: activation setup

OpenRouter provisions your account, credits, and invoicing. Your team confirms admin access through Settings > Preferences, and the shared channel opens. Once your organization is live, your team can create its own API keys independently through the Enterprise Quickstart, without waiting for the activation call. The call is where your FDE pairs with you on key creation and the rest of the configuration.

Day 1: activation working session

During the 60-minute activation call, your team will:
  1. Create your workspaces
  2. Set your guardrails and Zero Data Retention policy
  3. Set workspace budgets
  4. Start SSO domain verification
  5. Create a management key and your first API key
  6. Make a first test API call to confirm the key works
Create the management key from Management API Keys. If your organization uses provider credentials, review BYOK and API Key Rotation.

Day 2: first production inference

Your first real traffic runs from your own application environment. Usual blockers are outbound network restrictions, authentication, and environment setup, and your FDE works through them with you in the shared channel.

Onboarding: Day 3 to 14

Milestone: production traffic at your committed pace, signed off in writing by day 14.

Day 3: workspaces and budgets

Your team maps workspaces and sets budgets. The FDE walks your team through the structure. Start from your home dashboard, use Create Workspace, or automate setup with the Workspaces API.

Day 5: governance

Your team completes SSO, SCIM group mappings, guardrails, and data-retention policy. The FDE provides the DNS record and IdP guide. Review privacy settings, Data Collection, Provider Logging, and the Guardrails API as needed.

Day 5 to 7: inference strategy and observability session

The second and final required meeting. It covers provider routing, model fallbacks, prompt caching, presets so your application code names use cases rather than model versions, Broadcast into your own observability stack, user attribution, and access to the Analytics API (beta) for your own reporting. Create presets from the Presets page. You leave the session with at least one preset in use and traces flowing to your platform.

Day 7: observability

Your team configures Broadcast with the FDE pairing. Start from Settings > Observability, review traces in Logs or your workspace’s Observability settings, and compare Broadcast with Input & Output Logging.

Day 10: rollout

Your team rolls out beyond the first team. The FDE supplies SCIM group mappings and developer-ready setup material. Coding-agent teams can use Ori Harness for the rollout.

Day 14: go-live sign-off

Production traffic is holding. The FDE and CSM send the written sign-off summary through the shared channel.

Adoption: Day 15 to 30

Milestone: month-to-date usage at or above your commitment by day 30.

Day 15 to 20: production ramp

Your team brings the workloads you scoped into production. Usage, error rates, latency, and spend are reviewed against your baseline.

Day 21: expansion review

The written expansion review proposes a second workload drawn from your usage data.

Day 30: month-one review

The month-one report covers pace against commitment, spend by team and model, reliability, savings against direct provider access, and recommendations. Use Usage Accounting, Activity Export, and the Activity page for reporting.

Optimization: Ongoing

Monthly optimization cadence

Your team tunes provider routing, prompt caching, and presets during office hours and through the usage and optimization report. Review Uptime Optimization when tuning reliability and latency. The shared channel remains open between quarterly reviews.

Setting up well

Two documents cover the configuration decisions that are cheapest to get right early and most expensive to correct once your developers are active:

Enterprise Quickstart

Organization, workspaces, key management, security controls, presets, and observability, step by step.

Organization Management

Roles, shared credits, member management, and centralized usage tracking.
For application teams expanding beyond the first workload, see the Quickstart, Structured Outputs, Tool Calling, and Latency and Performance. For enterprise sales inquiries or custom requirements, contact our team at openrouter.ai/enterprise.