# Telnyx Inference Model Discovery

Use the live Models API to discover the models available to your account. This website does not maintain an authoritative model catalog.

Site index: https://telnyx.com/llms.txt

**Endpoint:** GET https://api.telnyx.com/v2/ai/openai/models
**Authentication:** your Telnyx API key as a Bearer token. No credentials belong in a public website URL.

```sh
curl --fail-with-body 'https://api.telnyx.com/v2/ai/openai/models' \
  -H "Authorization: Bearer $TELNYX_API_KEY"
```

Read the response's `data` array. Use IDs exactly as returned, including case. Inspect `pricing`, `context_length`, `max_completion_tokens`, `regions`, `regions_by_service_tier`, `recommended_for_assistants`, and `is_fine_tunable` instead of copying a static list.

With an OpenAI-compatible SDK, set the base URL to `https://api.telnyx.com/v2/ai/openai` and call `client.models.list()`.

Use the returned model IDs exactly as supplied. Results include models available to the authenticated account and may include private fine-tuned models; do not publish an account response as a global catalog. Assistant recommendations do not guarantee every inference model works for every assistant configuration.

`GET /v2/ai/models` is deprecated; new integrations should use `/v2/ai/openai/models`.

## Migrating from the website snapshot

The models, fine_tuning and regions fields are incomplete historical examples retained for compatibility. Do not use them for model selection, pricing, regional availability or fine-tuning decisions. Query model_discovery.url with your own Telnyx API key.

The JSON endpoint retains legacy field names and types until version 2.0, with explicit `_deprecations`. Its `updated_at` dates these discovery instructions, not a refresh of the old model/pricing snapshot. These discovery surfaces do not cache or proxy account credentials or API results.

[Machine-readable discovery instructions](/ai/inference-models.json) · [Schema migration notes](/ai/schema-changelog.json)