> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tensormesh.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# tm infer chat

> Call POST /v1/chat/completions on serverless inference.

## When To Use This

Use `tm infer chat` when you want to send an actual serverless inference request through the CLI.

## Usage

```sh theme={null}
tm infer chat [OPTIONS]
```

## Examples

### Send a non-streaming serverless request with an explicit inference API key.

```sh theme={null}
tm infer chat --api-key YOUR_INFERENCE_API_KEY --model YOUR_SERVERLESS_MODEL_NAME --json '[{"role":"user","content":"Say hello."}]'
```

### Stream tokens over SSE and fail a stalled stream after 30 seconds of upstream silence.

```sh theme={null}
echo '[{"role":"user","content":"Stream tokens."}]' | tm infer chat --api-key YOUR_INFERENCE_API_KEY --model YOUR_SERVERLESS_MODEL_NAME --stream --stream-idle-timeout 30
```

## Options

| Name                    | Type                 | Required | Default        | Details                                                                                              |
| ----------------------- | -------------------- | -------- | -------------- | ---------------------------------------------------------------------------------------------------- |
| `--surface`             | `choice[serverless]` | no       | `"serverless"` | Inference surface to use.                                                                            |
| `--model`               | `text`               | no       |                | Model name to use.                                                                                   |
| `--api-key`             | `text`               | no       |                | Inference API key (Authorization: Bearer ...).                                                       |
| `--base-url`            | `text`               | no       |                | Override the base URL for the selected surface.                                                      |
| `--stream`              | `boolean`            | no       | `false`        | Stream tokens via SSE. Boolean flag.                                                                 |
| `--stream-idle-timeout` | `float`              | no       | `300.0`        | Maximum idle read timeout in seconds for --stream responses.                                         |
| `--json`                | `text`               | no       |                | JSON payload or @file.json (object or messages array). When omitted, reads piped stdin if available. |
| `--file`                | `path`               | no       |                | Read JSON payload/messages from file.                                                                |
| `--timeout`             | `float`              | no       |                | HTTP connect timeout in seconds for the inference request.                                           |

## Inherited Global Options

| Name                  | Type           | Required | Default                              | Details                                                                                                                      |          |    |          |                                                                   |
| --------------------- | -------------- | -------- | ------------------------------------ | ---------------------------------------------------------------------------------------------------------------------------- | -------- | -- | -------- | ----------------------------------------------------------------- |
| `--version`, `-V`     | `boolean`      | no       | `false`                              | Show the version and exit. Boolean flag.                                                                                     |          |    |          |                                                                   |
| `--config`            | `path`         | no       | `"~/.config/tensormesh/config.toml"` | Path to config TOML file                                                                                                     |          |    |          |                                                                   |
| `--output`            | \`choice\[text | json     | yaml                                 | raw                                                                                                                          | table]\` | no | `"text"` | Output format (text is human-readable; json is machine-friendly). |
| `--quiet`             | `boolean`      | no       | `false`                              | Suppress non-essential output. Boolean flag.                                                                                 |          |    |          |                                                                   |
| `--debug`             | `boolean`      | no       | `false`                              | Print debug logs to stderr (secrets redacted). Boolean flag.                                                                 |          |    |          |                                                                   |
| `--ca-bundle`         | `path`         | no       |                                      | Path to a PEM CA bundle for TLS verification (overrides TENSORMESH\_CA\_BUNDLE).                                             |          |    |          |                                                                   |
| `--max-retries`       | `integer`      | no       |                                      | Max retries for idempotent HTTP requests on transient errors (overrides TENSORMESH\_MAX\_RETRIES; subcommands may override). |          |    |          |                                                                   |
| `--controlplane-base` | `text`         | no       |                                      | Override the Control Plane base URL.                                                                                         |          |    |          |                                                                   |

## Auth Scope

* inference-api-key

## Prerequisites

* Provide `--model` explicitly or include `model` in the JSON request body.
* If you have Control Plane access for the same Tensormesh environment, discover published serverless model names with `tm billing pricing serverless list`, then pass the returned `pricing[].model` value with `--model`.
* If you only have inference credentials, or you are targeting a different serverless host override, get the model name from your Tensormesh environment before sending the request.

## Caveats

* Reuses `gateway_api_key` from `config.toml` as the inference API key when `--api-key` is omitted.
* `gateway_api_key` is the stored inference API key used by the SDK as `inference_api_key`.
* `tm billing pricing serverless list` helps discover published serverless model names for the current Tensormesh Control Plane environment. If you are targeting a different serverless host override, confirm the model name for that host separately.
* `--stream` currently supports only `--output text`.

## Related Commands

* [`tm billing pricing serverless list`](/cli/reference/billing/pricing/serverless/list)
* `tm models list`
* [`tm doctor`](/cli/reference/doctor)

## Parent Command

* [`tm infer`](/cli/reference/infer)