Skip to main content

​
Before You Use tm

  • Install the CLI first if tm is not already on your PATH. In this repo checkout, use ./.venv/bin/tm until your shell already exposes tm.
  • Start with tm init if you are not sure what is already configured locally.
  • Use tm auth login followed by tm init --sync for the standard persistent On-Demand setup flow.
  • Put global options such as --output json before the subcommand.
  • Set TM_CONFIG_HOME when you need an alternate config and auth root for local scripting.
  • Use tm auth status, tm infer doctor, and tm doctor for local readiness checks.
  • Use tm auth whoami for a live Control Plane auth check and tm infer chat for an end-to-end inference check.

​
Usage

tm [OPTIONS] COMMAND [ARGS]...

​
Examples

​
Inspect local setup and print the shortest path to a working request.

tm init

​
Inspect gateway readiness before the first request.

tm infer doctor

​
Inspect local Control Plane and inference auth state.

tm auth status

​
Send a basic chat request after the standard On-Demand setup flow.

tm infer chat --json '[{"role":"user","content":"hi"}]'

​
Details

Tensormesh CLI for Control Plane and Inference API workflows.

​
Subcommands

  • tm activities: Activities (Control Plane).
  • tm admin: Admin-only commands (Control Plane).
  • tm auth: Authenticate to the Tensormesh Control Plane.
  • tm billing: Billing and usage (Control Plane).
  • tm completion: Print a shell completion script for Bash/Zsh/Fish.
  • tm config: Manage Tensormesh CLI configuration.
  • tm doctor: Diagnose configuration and credentials.
  • tm infer: Inference commands for shared serverless and On-Demand OpenAI-compatible endpoints.
  • tm init: Guide first-time CLI setup for Control Plane and Gateway workflows.
  • tm logs: Fetch deployment logs (GET /v1/observability/deployments/{deploymentId}/logs).
  • tm metrics: Observability metrics (Control Plane).
  • tm models: Control Plane model resources.
  • tm products: Products (Control Plane).
  • tm reserved-deployments: Reserved deployments (Support; Control Plane).
  • tm tickets: Support tickets (Control Plane).
  • tm users: User profile resources (Control Plane).
  • tm version: Show CLI version and runtime details.

​
Options

NameTypeRequiredDefaultDetails
--version, -VbooleannofalseShow the version and exit. Boolean flag.
--configpathno"~/.config/tensormesh/config.toml"Path to config TOML file
--output`choice[textjsonyamlrawtable]`no"text"Output format (text is human-readable; json is machine-friendly).
--quietbooleannofalseSuppress non-essential output. Boolean flag.
--debugbooleannofalsePrint debug logs to stderr (secrets redacted). Boolean flag.
--ca-bundlepathnoPath to a PEM CA bundle for TLS verification (overrides TENSORMESH_CA_BUNDLE).
--timeoutfloatnoDefault HTTP timeout in seconds (overrides TENSORMESH_TIMEOUT_SECONDS; subcommands may override).
--max-retriesintegernoMax retries for idempotent HTTP requests on transient errors (overrides TENSORMESH_MAX_RETRIES; subcommands may override).
--controlplane-basetextnoOverride the Control Plane base URL.
--gateway-providertextnoInference Gateway provider for built-in host selection (nebius, lambda, yotta).

​
Environment Variables

  • TENSORMESH_CA_BUNDLE
  • TENSORMESH_TIMEOUT_SECONDS
  • TENSORMESH_MAX_RETRIES
  • TM_CONFIG_HOME

​
Additional Notes

Config precedence: flags > env > config.toml ([overrides] before [managed]) > defaults. Standard persistent setup: tm auth login tm init —sync

​
updates [managed] in the active config.toml state root

Global options such as —output must come before the subcommand: tm —output json auth status Gateway credentials (gateway_api_key, gateway_user_id, gateway_model_id) have no environment variable equivalents. Set them in [overrides] in config.toml or use tm init —sync to sync [managed]. Control Plane auth uses auth.json under the active state root; inference commands normally read gateway credentials from [managed]. Quickstart: tm auth login tm init —sync tm auth status tm infer chat —json ’[{“role”:“user”,“content”:“hi”}]’