Use Cases
Three real scenarios. One gateway.
Use case
You're building with Claude Code, LangChain, CrewAI, or your own agent framework. A single task spawns an orchestrator, multiple subagents, and dozens of tool calls across different models. At the end of the day, you have no idea what anything cost.
Wardis correlates every request back to the task that triggered it. One dashboard shows you what each agent task actually cost — broken down by orchestrator, subagent, and model.
An agent team with 5 subagents runs for 20 minutes. Your provider dashboard shows 247 individual API calls. Which task caused them? What did each subagent cost? Was the orchestrator's model choice efficient?
Without task-level correlation, you're looking at a flat list of requests with no structure. You can see that you spent $4.72 but not why.
Add a header to your agent requests — or use the built-in MCP server for Claude Code. Wardis groups all 247 requests into one task, builds the agent tree, and shows you cost at every level.
The Agents dashboard shows total tasks, cost by agent type, and a timeline. Click any task to see the full tree with per-agent cost breakdown.
Use case
You build a SaaS product that uses AI under the hood. Your customers don't know or care about LLM providers — they use your app. But you need to know exactly how much AI each customer consumes so you can price, budget, and alert accordingly.
Wardis does this out of the box with zero code changes to your app. Map teams to customers, assign a key per customer, set per-customer budgets. Your app just sends the right API key per customer — Wardis handles the rest.
Use case
Your organization has multiple teams — engineering, data science, marketing, support — all using AI. Each team has its own budget, its own models, and its own usage patterns. Finance needs a single view of total spend. Security needs an audit trail. Engineering needs the flexibility to ship fast.
Wardis gives each team autonomy within their budget while giving leadership full visibility and control across the organization.
One dashboard with total AI spend, cost by team, cost by provider, and trend over time. Budget enforcement ensures no team can overspend. Cost Intelligence identifies optimization opportunities.
Team-scoped API keys and budgets. Developers create their own keys, see their own usage. Smart routing picks the right model for the task. Semantic cache reduces duplicate calls. No bottlenecks.
RBAC with four roles controls who can do what. Audit log tracks every management action. SSO/SAML for centralized access. All data stays on your infrastructure — self-hosted by design.
Self-host Wardis in minutes. Free and open source.