Introducing Verdictan:
One gateway between your apps and 60+ AI providers. Enforce policy, control spend, redact PII, and keep a complete audit trail — with a single YAML config.
curl -fsSL https://verdictan.com/install | shAvailable for macOS, Linux and Windows. Read the setup documentation
Explore AI provider model catalogs
Apply policy to supported agent traffic
A configured gateway evaluates its selected policy chain for supported agent requests. It can record decision events for later review.
Read the agent governance guide
Reuse verified work in the correct scope
Context Fabric searches authorized team, repository and branch scopes. Agents can use verified prior work without treating every earlier answer as trusted.
Read the Context Fabric guide
Understand estimated AI costs
Review token use and estimated costs by model, agent, team or API key. Use the provider invoice as the authoritative cost record.
Read about usage costs
Review configured budget controls
Set thresholds and declared hard limits for supported scopes. Verify the active request path before you rely on a gateway block.
Read about budget controls
FAQ
What is Verdictan?
Verdictan is a self-hosted AI governance gateway. It routes supported AI traffic through configured policy, provider and evidence controls.
How do I use Verdictan?
Install the Verdictan CLI, create policy-config.yaml, point a supported client at the gateway and verify the matching decision event before you expand the rollout.
Which AI traffic can Verdictan govern?
Verdictan documents policy enforcement for supported Chat Completions, Responses and Messages traffic. Check the request-family guide before you route other traffic.
Which AI providers does Verdictan support?
The public model catalog covers more than 60 AI providers. Check each provider page for its available models and routing details.
Can I self-host Verdictan?
Yes. Verdictan supports self-hosted gateway deployment profiles for cloud, on-premises and disconnected environments.
Does Verdictan replace my provider invoice?
No. Verdictan reports estimated token use and costs for supported traffic. Your provider invoice remains the authoritative cost record.