One API across many models
Route OpenAI-compatible requests through a single endpoint while choosing from hundreds of models and many providers.
TrustedRouter
TrustedRouter is an OpenAI-compatible AI router for hundreds of models through one API, with verifiable privacy, provider failover, and BYOK support.
What TrustedRouter does
TrustedRouter is an OpenAI-compatible AI router and gateway for teams that want to access many models through one API while keeping the prompt path verifiable. The product is positioned as a privacy-first alternative to OpenRouter, with attested infrastructure, no prompt or output logs, and routing options that make privacy, geography, cost, and failover explicit.
It supports more than 600 models and dozens of providers, with named aliases for common routing goals such as healthy-provider rollover, zero-retention routing, confidential workloads, and Europe-focused routes. The site also emphasizes easy migration: existing OpenAI SDK calls can keep working after a base URL change.
Core capabilities
Policy-based routing aliases
Use aliases such as trustedrouter/auto, trustedrouter/zdr, trustedrouter/e2e, and trustedrouter/eu to make routing intent explicit.
Attested gateway verification
Keep traffic on a verifiable prompt path with an attested gateway, source commit and image digest checks, and fail-closed behavior if attestation does not pass.
Provider failover and regional routing
Shift traffic to healthy providers and regional routes when an upstream service is unavailable, instead of relying on one provider or one cloud.
BYOK support
Bring your own provider keys where needed while still using the TrustedRouter routing layer and trust boundary.
Public catalog and no content logs
Review catalog pricing, route metadata, and public status before sending traffic, with no prompt or output content stored in the control plane.
Common ways teams use TrustedRouter
Migrate an existing OpenAI integration
Teams can keep their current OpenAI SDK integration and point it at TrustedRouter to reach a wider model catalog without rewriting their application code.
Add provider failover to production traffic
Applications that need a simple production default can use trustedrouter/auto to roll over to a healthy provider when one upstream route degrades or fails.
Handle sensitive or confidential prompts
Legal, compliance, or other sensitive workflows can route through trustedrouter/zdr or trustedrouter/e2e to match stricter privacy requirements.
Route workloads by geography
Teams working across regions can choose trustedrouter/eu when they want Europe-focused routing rather than a generic global route.
Preserve existing provider agreements
Organizations with existing provider commitments can use BYOK to keep their rate limits or negotiated spend while still using TrustedRouter’s routing layer.
Pros and Cons
Pros
- Works with the OpenAI SDK and existing OpenAI-shaped calls after a base URL change.
- Offers a large public catalog of models and provider routes with pricing and route metadata.
- Provides attested gateway verification, with source and digest checks described on the site.
- Includes routing aliases for privacy, region, and failover needs, including zero-retention and Europe-focused options.
- Supports BYOK and usage-based billing without a monthly subscription.
Cons
- The source does not show a free tier; pricing is usage-based with prepaid credits or BYOK.
- Some details are policy- and provider-dependent, so exact behavior can vary by route or model.
- The product is developer-oriented and appears to require comfort with API configuration rather than a no-code interface.
FAQ
What is TrustedRouter?
TrustedRouter is an OpenAI-compatible AI gateway that routes requests across many model providers through one API. It emphasizes verifiable privacy by separating the prompt path from the dashboard and billing control plane.
Can I keep using my existing OpenAI client?
Yes. The site says you can keep the OpenAI SDK and change only the base URL to point at TrustedRouter, then continue using chat completions or Responses-style calls.
How do routing aliases work?
TrustedRouter offers routing aliases such as trustedrouter/auto for healthy-provider rollover, trustedrouter/zdr for zero-retention routing, trustedrouter/e2e for confidential and E2EE routes, and trustedrouter/eu for Europe-focused routing.
Does TrustedRouter log prompts or outputs?
No. The site states that prompt and output content are intentionally excluded from the control plane and metadata systems, with optional content export only when explicitly configured at the destination level.
How does pricing work?
Pricing is usage-based. The pricing page says text and embedding routes are billed at provider cost plus 5.5%, with no monthly plan, and BYOK is supported for providers where you already have committed spend or enterprise rate limits.
Quick Facts
- Category
- AI infrastructure / developer tool
- Primary use
- OpenAI-compatible routing across many model providers
- API
- OpenAI-compatible
- Pricing model
- Usage-based; prepaid credits or BYOK; no monthly subscription
- Source domain
- trustedrouter.com
- Notable workflow
- Change base_url and keep the existing OpenAI client
TrustedRouter Alternatives
AakarDev AI
AakarDev AI helps teams manage AI provider access, project setup, logs, and analytics in one dashboard. BYOK support included.
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repos and verifies changes with compilers, debuggers, sanitizers, and tests.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads in Firecracker micro-VMs with private networking and SDK, CLI, or MCP control.
hob
hob is an independent workspace for coding agents, with local control over sessions, terminals, history, routing, and follow-up work.
Ably Chat
Ably Chat is a chat API platform for custom realtime chat apps, with rooms, typing indicators, presence, reactions, message updates and usage-based pricing.
Manta AI
Manta AI is an autonomous web app testing tool that maps app behavior, catches regressions, and generates tests from a URL, no scripts or selectors needed.