---
title: "TIR GenAI API FAQ"
description: "Short answers about TIR GenAI API model availability, regions, authentication, and billing, with links to the full procedure."
content_type: "faq"
product: "genai-api"
platform: "TIR"
url: "/tir/genai-api/faq/"
---

# TIR GenAI API FAQ

For step-by-step fixes, see [Troubleshooting TIR GenAI API](/tir/genai-api/troubleshooting/genai-api/).

## Models and availability

### Is a model on GenAI API guaranteed to stay available?

No. E2E can mark any model out of inventory or remove it from the catalog, including a model used
in the examples in these docs. Check a model's own page before
you rely on it, and see [what happens when a model is
unavailable](/tir/genai-api/troubleshooting/genai-api/#a-model-shows-request-instead-of-opening-its-playground).

### What happens when I request a model that is unavailable?

**Request** files a support ticket for you, and you get an email when the model is available. A
**Benchmarking Request** also files a ticket, and you get an email when the results are ready.
Neither request changes the model automatically.

### Is GenAI API available in both Delhi and Chennai?

Yes. GenAI API shows the same model catalog in both regions.

## Authentication

### Does my Personal Access Token expire?

Yes. A token and its Auth Token stop working 365 days after you create it. Create a new one from
**Personal Access Tokens** when it expires. See [Get started with TIR GenAI
API](/tir/genai-api/get-started/).

### Can I have more than one Personal Access Token?

Yes. Create as many as you need, and delete any one of them from the same screen. To replace a
token, delete it and create a new one.

## Calling the API

### Which calling method should I use for a model?

It depends on the model's endpoint. A text or embedding model has its name in the endpoint URL,
so call it with the OpenAI SDK or a plain REST request. A speech, image, or translation model takes
its name in the request body, so call it with the TIR SDK or REST. See [Models and calling
methods](/tir/genai-api/concepts/models-and-calling-methods/).

### Does Max Tokens limit the model's memory of the conversation?

No. **Max Tokens** caps the length of one reply. How much of the conversation the model can take in
is its context length. See [My reply stops
partway through](/tir/genai-api/troubleshooting/genai-api/#my-reply-stops-partway-through) for how
a model's context length works.

## Bulk prompts

### Is there a limit on how many prompts I can upload in one bulk prompt file?

No published limit applies to how many prompts a file holds or how many bulk jobs run at once. You
upload one file at a time. See [Run bulk prompts](/tir/genai-api/how-to/run-bulk-prompts/).

## Billing

### Is GenAI API billed per request?

No. GenAI API bills for what each request uses, such as its input and output tokens, rather than a
flat amount per request. See [Billing and
usage](/tir/genai-api/concepts/billing-and-usage/).

### Is there a setup charge or a commitment to use GenAI API?

No. You pay only for what you call, with no setup charge, reserved capacity, or long-term
commitment. For prices, see [E2E Networks pricing](https://www.e2enetworks.com/pricing).