> ## 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.

# First Inference Request

> Send a first chat request through the Tensormesh Serverless inference flow.

This page covers the standard first-request flow against the serverless inference surface.

If `tm` is not already on your `PATH`, install the CLI first with
[Installation](/cli/guides/installation). The commands below assume `tm` is on
your `PATH`. If you are running from this repo checkout without activating a
shell that already exposes `tm`, use `./.venv/bin/tm`.

The shortest serverless request:

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

If you do not already know a valid serverless model name, use [`tm billing pricing serverless list`](/cli/reference/billing/pricing/serverless/list) before using that shortcut.

Use `tm infer doctor` when you want to check whether the local config and credentials are wired up correctly before sending the request.

## Get An API Key

If your environment exposes self-serve key creation, use the authenticated Control Plane flow:

```sh theme={null}
tm auth login
USER_ID="$(tm --output json auth whoami | python3 -c 'import json,sys; print(json.load(sys.stdin)["user"]["id"])')"
tm users api-keys create --user-id "$USER_ID" --name cli-key --yes
```

Otherwise, ask your operator or admin for the exact inference API key to use.

## Send A Chat Request

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

For streaming output:

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

Streaming requests use a bounded idle read timeout. If you expect long quiet gaps between SSE events, raise it explicitly:

```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 600
```

## Related Reference

* [`tm infer`](/cli/reference/infer)
* [`tm infer doctor`](/cli/reference/infer/doctor)
* [`tm infer chat`](/cli/reference/infer/chat)
* [`tm init`](/cli/reference/init)
* [`tm billing pricing serverless list`](/cli/reference/billing/pricing/serverless/list)
* [API Quickstart](/api-quickstart) for raw HTTP equivalents of these flows