# Telnyx AI: AI Gateway — Documentation Index
> AI Gateway documentation within the AI section of the Telnyx developer docs (https://developers.telnyx.com).
> This file: https://developers.telnyx.com/docs/development/llms/ai-ai-gateway-llms-txt · Root index: https://developers.telnyx.com/llms.txt

## AI Gateway

- [Overview](https://developers.telnyx.com/docs/inference/ai-gateway.md): AI Gateway issues scoped inference credentials for applications: token groups and keys, model allowlists, budgets, rate limits and usage ledgers.
- [Quickstart](https://developers.telnyx.com/docs/inference/ai-gateway/quickstart.md): Create a token group, issue a scoped token key, call a model with the OpenAI SDK, inspect usage and revoke the key.

## API

- [Inference API](https://developers.telnyx.com/docs/inference/ai-gateway/inference-api.md): Call models through AI Gateway with the OpenAI and Anthropic SDKs: base URL, authentication, streaming, request limits and supported request fields.
- [Management API](https://developers.telnyx.com/docs/inference/ai-gateway/management-api.md): Provision AI Gateway token groups, users, keys, end-user caps and provider keys, with idempotency, ETag preconditions and snapshot pagination.

## Concepts

- [Budgets & rate limits](https://developers.telnyx.com/docs/inference/ai-gateway/controls.md): How AI Gateway enforces budgets, rate limits and end-user identity, and what an application observes at each limit.
- [Guardrails](https://developers.telnyx.com/docs/inference/ai-gateway/guardrails.md): Configure token-group policies to flag or block supported secrets and sensitive data in prompts and model responses, including streamed responses.
- [Bring your own key](https://developers.telnyx.com/docs/inference/ai-gateway/byok.md): Attach an OpenAI or Anthropic key to AI Gateway groups so your provider bills inference while Telnyx enforces budgets, rate limits and attribution.
- [Usage reporting](https://developers.telnyx.com/docs/inference/ai-gateway/usage.md): Query AI Gateway spend events and dimensional summaries by group, user, key or end user, and interpret cost and unknown-usage fields correctly.
- [Errors](https://developers.telnyx.com/docs/inference/ai-gateway/errors.md): HTTP status codes, structured error codes and error envelopes returned by the AI Gateway management and inference planes, with handling guidance.