# uip llm-gateway

> Syntax and options for `uip llm-gateway models`, which lists LLM models available to a product/feature via the UiPath LLM Gateway discovery API.

`uip llm-gateway` lists the LLM models available via the **UiPath LLM Gateway** — the platform-managed routing layer that resolves which underlying model (Claude, GPT, etc.) backs a given product feature, accounting for tenant policy, BYO model configuration, and preview gating. There is one command: `models`.

See [Tools (plugins)](./concepts-tools.md) for how this tool integrates with the CLI. For managing BYO LLM connections and guardrail configurations that this discovery result reflects, see [`uip llm-configuration`](./uip-llm-configuration.md) and [`uip guardrails`](./uip-guardrails.md).

## Synopsis

```
uip llm-gateway models --product <name> --feature <name> [--operation <code>] [--login-validity <minutes>]
```

## uip llm-gateway models

List LLM models available for a given product and feature scope, as resolved by the LLM Gateway discovery API — accounting for tenant policy blocks, preview flags, and BYO subscriptions.

### Options

| Long | Value | Required | Description |
|---|---|---|---|
| `--product <name>` | string | **yes** | Requesting product scope, e.g. `guardrails`. |
| `--feature <name>` | string | **yes** | Requesting feature scope, e.g. `llm-as-judge-validator`. |
| `--operation <code>` | string | no | Narrow results to a specific operation code, e.g. `Guardrails.LlmAsJudge.Debug`. |
| `--login-validity <minutes>` | integer | no | Override the interactive-login token lifetime for this call. |

### Examples

```bash
# List LLM models available for the LLM-as-judge guardrail
uip llm-gateway models --product guardrails --feature llm-as-judge-validator
```

```bash
# List models scoped to the debug operation mode
uip llm-gateway models --product guardrails --feature llm-as-judge-validator \
  --operation Guardrails.LlmAsJudge.Debug
```

```bash
# List models for a different product and feature
uip llm-gateway models --product agenthub --feature agent-health-score-linter
```

### Data shape (--output json)

```json
{
  "Code": "LlmGatewayModelsList",
  "Data": [
    {
      "ModelId": "anthropic.claude-haiku-4-5-20251001-v1:0",
      "DisplayName": "Claude Haiku 4.5",
      "Vendor": "AwsBedrock",
      "ModelFamily": "AnthropicClaude",
      "SubscriptionType": "UiPathOwned",
      "IsByo": false,
      "IsPreview": false,
      "IsBlockedByPolicy": false
    }
  ]
}
```

`IsByo` is `true` when the model comes from a bring-your-own connection configured via [`uip llm-configuration byo-connections`](./uip-llm-configuration.md). `IsBlockedByPolicy` reflects tenant-level governance policy, independent of whether the model would otherwise be available. An empty `Data` array means no models are available for that product/feature combination — not an error.

## Error behavior

Any failure (including no active login) returns `Failure` with `Instructions: "Run 'uip login' to authenticate before using this command."` — this is the CLI's generic fallback instruction here, not necessarily the true cause; check the accompanying `Message` for the underlying error.

## See also

- [`uip llm-configuration`](./uip-llm-configuration.md) — manage BYO LLM connections that this discovery result can reflect.
- [`uip guardrails`](./uip-guardrails.md) — a consumer of LLM Gateway model discovery for guardrail configuration.
- [Tools (plugins)](./concepts-tools.md)
- [Global options](./global-options.md)
- [Exit codes](./exit-codes.md)