OpenAI builds and deploys advanced AI models like GPT-4o for autonomous agents and workflows.
anonrouter
$0.75Starting price
0Popularity
About anonrouter
anonrouter provides a private gateway for routing requests to AI models while maintaining confidentiality. It offers a single OpenAI-compatible endpoint that connects to over 200 models across multiple providers. Requests are processed inside an attested Intel TDX enclave, ensuring prompts and responses remain sealed and unreadable by the service. Usage is metered only from token counts and request metadata, with no prompts stored or used for training. The platform includes workspace controls for managing API keys, routing behavior, and privacy settings without exposing content. It supports both text and media models through a unified API, with automatic routing to the cheapest capable model within privacy boundaries. The service operates on a pay-as-you-go model with published provider rates and an 8% service fee, with no monthly charges or prompt logs retained.
Key features
- OpenAI-compatible API endpoint
- Confidential compute with Intel TDX enclave
- Zero prompt retention and no training data usage
- Automatic model routing and smart fallbacks
- Workspace controls for API key management
- Support for text and media models
- Usage metering by token counts only
- Opaque sessions and hashed API keys
Use cases
- Private AI inference for enterprises requiring confidentiality
- Unified API for multiple AI model providers
- Secure agent and IDE integration with private routing
Pros
- Single OpenAI-compatible endpoint for 200+ models
- Prompts processed in attested Intel TDX enclave with no retention
- Workspace controls for API key management and routing configuration
- Supports both text and media models through unified API
- Pay-as-you-go pricing with published provider rates
Cons
- No free tier available
- Requires credit purchase for usage
- Limited to OpenAI API compatibility
- Rate limits apply (30 req/min, 250K tokens/min)
Frequently asked questions about anonrouter
What is anonrouter and what does it do?
anonrouter is a private AI routing service that provides a single OpenAI-compatible endpoint to access over 200 models across multiple providers while maintaining confidentiality. It processes requests inside an attested Intel TDX enclave, ensuring prompts and responses remain sealed and unreadable by the service.
Who should use anonrouter?
The tool is designed for developers, teams, and organizations that require private AI inference without exposing prompts or responses. It suits use cases where confidentiality is critical, such as handling sensitive data or proprietary workflows.
How does anonrouter ensure privacy?
Requests are processed inside an Intel TDX enclave, which seals prompts and responses so they cannot be read by the service. Usage is metered only from token counts and request metadata, with no prompts stored or used for training.
Does anonrouter support media models?
Yes, anonrouter supports both text and media models through a unified API, allowing users to generate images and speech from various media-capable models while maintaining privacy.
How do I get started with anonrouter?
Users can create an account using a recovery phrase and generate an API key. The service can then be integrated by pointing any OpenAI-compatible client, agent, or IDE to the anonrouter endpoint, such as `https://api.anonrouter.ai/v1`.
What pricing model does anonrouter use?
anonrouter operates on a pay-as-you-go model, with usage metered from token counts and request metadata. It includes a service fee on top of published provider rates, with no monthly charges or prompt logs retained.
