TrustedRouter

What is TrustedRouter?

TrustedRouter is a privacy-first AI gateway that puts over 600 models from over 100 providers behind one OpenAI-compatible API. Prompts and outputs pass through an attested gateway that stores no content. Teams switch by changing a single base URL, then pick aliases for failover, zero retention, confidential compute, or Europe-only routes.

Top Features:

• One API: reach over 600 models through the OpenAI SDK and one URL.
• Privacy routing: named aliases send traffic to zero-retention, confidential, or Europe-focused providers.
• Verifiable trust: an attested gateway publishes evidence that prompts and outputs are never stored.

Use Cases:

• Private inference: send sensitive prompts only to providers with zero data retention.
• Provider failover: keep apps running by shifting traffic to healthy providers during outages.
• Gateway migration: move existing OpenRouter workloads over by swapping the base URL.

Who Can Use TrustedRouter?

• Developers: build apps on many models without juggling separate provider keys and SDKs.
• Security teams: check attestation evidence before approving AI traffic from company apps.
• European companies: keep model requests on Europe-focused routes to support data residency needs.

Pricing

• Text and embeddings (provider cost plus 5.5%): pay as you go with no subscription or monthly fee.
• Video generation (provider quote plus 20%): billed per request on top of the direct provider price.
• Migration credits (by approval): available to teams spending over 100 dollars monthly on language models.

Pros and Cons

Pros:

• No content logging: prompts and outputs stay ephemeral instead of sitting in logs.
• Open source: the control plane, backend, SDKs, and infrastructure code are all public.
• Flexible payments: card, ACH, PayPal, and stablecoin all work for prepaid credits.

Cons:

• Usage markup: every text request carries a 5.5 percent fee above provider cost.
• No free tier: buy credits or bring your own keys before sending any requests.
• Developer focus: setup assumes API and SDK knowledge, so newcomers may struggle.

FAQs:

1) Does it work with the OpenAI SDK?

Yes, keep the OpenAI SDK and just change the base URL.

2) Does it store my prompts?

No, the gateway keeps prompts and outputs ephemeral and never logs them.

3) How many models can I reach?

Over 600 models from more than 100 providers, all through one key.

4) Is there a subscription fee?

No, you pay provider cost plus a small markup, nothing monthly.

5) Can I use my own provider keys?

Yes, bring-your-own-key billing works alongside prepaid credits and usage billing.