Frequently Asked Questions

Exerati is an AI infrastructure platform for developers and businesses. We provide an OpenAI-compatible API, CLI access, and a UI-to-Code and Vision model testing platform. You can access multiple AI models through a single, unified interface.

Exerati is designed for developers, engineering teams, and businesses that need reliable AI model access. Whether you are building a chatbot, analyzing images, generating code, or prototyping AI features, Exerati provides the infrastructure you need.

We provide access to a curated catalog of models across text, vision, code, and audio modalities. Browse our Models page for the full list with pricing and capabilities. New models are added regularly.

Yes. Our API follows the OpenAI specification. You can use existing OpenAI SDKs by changing the base URL to https://api.exerati.com/v1. Request and response formats are identical.

We do not currently offer proprietary SDKs. However, because our API is OpenAI-compatible, you can use the official OpenAI Python, Node.js, and other community SDKs with a simple base URL change. Proprietary SDKs are on our roadmap.

Default rate limits are 60 requests per minute per API key. Premium accounts get 600 requests per minute, and enterprise accounts get up to 6,000. You can configure individual key limits from your dashboard.

Yes. Our CLI tool allows you to connect to AI models directly from the terminal. You can pipe data, chain commands, and integrate into CI/CD pipelines and scripts.

Exerati uses a wallet-based billing system. You add funds to your wallet, and API usage is deducted per token based on the model pricing. There are no subscriptions or hidden fees — you only pay for what you use.

Yes. Every new account receives $10 in free test credit. You can use this to explore models, test integrations, and build prototypes without any commitment.

Navigate to the Wallet section in your dashboard and click Top Up. We currently support simulated payments for beta testing. Real payment processing via Stripe is coming soon.

Yes. Each API key can have a budget cap to prevent unexpected charges. When the cap is reached, requests using that key will be rejected until the cap is increased or reset.

API keys are hashed with bcrypt before storage. We only store the first 8 characters (key prefix) for identification purposes. The full key is shown exactly once at creation and cannot be retrieved later.

We log API request metadata (endpoint, model, token counts, timing) for usage tracking and debugging. We do not log request or response payloads. All logs are immutable and retained for 90 days.

Yes. Two-factor authentication is available for all accounts. Enable it from your Profile settings for an additional layer of security.