Models

Register, upload, and inspect models and adapters for dedicated model inference from your terminal.

Register, upload, and inspect the models and adapters you deploy with dedicated model inference. Commands run within a Together project.

Bringing your own weights is a two-step flow:

1. create registers a model record (metadata only).
2. upload or remote-uploads create adds the weight files.

To deploy a registered model, run tg endpoints deploy or its alias, tg models deploy. For the end-to-end workflow, see Custom models.

These commands require Together CLI 3.0 or later. In CLI 2.x they were tg beta models, and tg models listed the serverless catalog and uploaded your own models for the retired v1 dedicated endpoints.

Create

Register a new model record. This creates metadata only. It does not upload any files. Upload the weights afterward with upload or remote-uploads create.

Shell

tg models create my-custom-glm \
  --base-model zai-org/GLM-5.2

You can also pass a base model ID (ml_...). Values that do not start with ml_ are treated as a deploy model name: the CLI searches public supported models and requires an exact match on a deployment profile’s modelName. If the name is missing or matches more than one profile, the CLI prints the candidate model names and ids and asks you to re-run with an exact name or ml_... id.

Alias: tg models -c.

Parameters

Flag	Description
NAME	(required) The inference-addressable name for the model.
--type [model | adapter]	Whether the record holds full weights (model) or a LoRA adapter (adapter). Set once at create time. File operations derive the type from the record. Default: model.
--base-model [string]	(required) The supported base model this model derives from. Accepts a base model ID (ml_...) or an exact deployment profile modelName (for example Qwen/Qwen3.5-9B-FP8). Run tg models public to find either value.

Upload

Upload weight files from your local machine to an existing model or adapter record.

Shell

tg models upload ml_abc123 ./my-model-weights

Parameters

Flag	Description
MODEL-ID	(required) The existing model or adapter ID to upload files to.
LOCAL-PATH	(required) The local file or directory to upload.

List

List the models in the current project.

Shell

tg models list

Alias: tg models ls.

Parameters

Flag	Description
--limit [number]	Maximum number of models to return.
--after [string]	Pagination cursor to start from.

Retrieve

Get a single model by ID.

Shell

tg models retrieve ml_abc123

Alias: tg models get.

Update

Update a model record. Only the fields you pass are changed.

Shell

tg models update ml_abc123 \
  --name my-renamed-model \
  --description "Fine-tuned GLM for support triage"

Parameters

Flag	Description
ID	(required) The model ID to update.
--name [string]	New inference-addressable name.
--description [string]	New description.

Delete

Delete a model record. This removes only the metadata, not the uploaded files.

Shell

tg models rm ml_abc123

Alias: tg models -d.

Configs

List the deployable configs for a model. Pass a config ID to tg endpoints deploy --config to deploy that specific configuration.

Shell

tg models configs ml_abc123
tg models configs projects/proj_abc123/models/ml_abc123
tg models configs Qwen/Qwen3.5-9B

Pass a bare model ID (ml_...), a resource path (projects/{project_id}/models/{model_id}), or a model name. Names with a / resolve through the private models list when the prefix matches your project slug, or through public supported models otherwise. Ambiguous public names print a short error that points you at tg models public --search.

Parameters

Flag	Description
MODEL	(required) The model to list configs for. Accepts a model ID (ml_...), a resource path, or a model name.
--limit [number]	Maximum number of configs to return.
--after [string]	Pagination cursor to start from.

List files

List the files in a model or adapter.

Shell

tg models ls-files ml_abc123

Parameters

Flag	Description
ID	(required) The model or adapter ID.
--revision [string]	Revision ID to filter files for.

List revisions

List the revisions of a model, including each revision’s validation status.

Shell

tg models ls-revisions ml_abc123

Alias: tg models list-revisions.

Parameters

Flag	Description
ID	(required) The model ID.

Download

Download the weight files of a model or adapter to a local directory.

Shell

tg models download ml_abc123 ./local-dir

Parameters

Flag	Description
MODEL-ID	(required) The model or adapter ID (a project/name path or object ID).
LOCAL-PATH	(required) The local directory to write files into.
--revision [string]	Pin the download to a specific revision. Defaults to the latest.
--files [string]	Restrict to specific file paths. Repeatable, and commas are allowed.
--format [hf]	Output layout. Use hf for a Hugging Face snapshot layout.

List public models

List the publicly visible models across all projects. Use this to find a base model ID or deployment profile modelName for create and tg endpoints deploy.

Shell

tg models public \
  --product dedicated \
  --modality text

The human-readable table prints one row per deployment profile, with columns for Model (the profile’s modelName, such as Qwen/Qwen3.5-9B-FP8), GPUs, and Parallelism. Pass that Model value as the MODEL argument to tg endpoints deploy when you want that specific profile.

Add --json to see each architecture’s full record, including its id (which subsequent operations require) and its deploymentProfiles array. A model can expose more than one profile, each pairing a certified config with a hardware and parallelism choice, so read the profile’s modelName, config, gpuType, and gpuCount before deploying. See Choose a hardware config.

Parameters

Flag	Description
SEARCH	Search by ID, name, or description.
--limit [number]	Maximum number of models to return.
--after [string]	Pagination cursor to start from.
--modality [text | image | audio | video]	Filter by input modality.
--product [serverless | dedicated | fine-tuning]	Filter by product surface.

List org models

List the internal-visibility models in your organization.

Shell

tg models org

Parameters

Flag	Description
--limit [number]	Maximum number of models to return.
--after [string]	Pagination cursor to start from.

Remote uploads

Import model weights server-side from a Hugging Face repo or a presigned S3/GCS URL, without downloading them locally first. Create a remote upload job against an existing model record, then poll it until it succeeds.

Remote uploads create

Start a remote upload job.

Shell

tg models remote-uploads create ml_abc123 \
  --from https://huggingface.co/zai-org/GLM-5.2

For gated or private Hugging Face repos, pass your Hugging Face token with --token. For S3 or GCS, pass a presigned archive URL as --from (no token needed).

Alias: tg models remote-uploads -c.

Parameters

Flag	Description
MODEL-ID	(required) The existing model or adapter ID to upload files to.
--from [string]	(required) The remote source URL (a Hugging Face repo or a presigned S3/GCS URL).
--token [string]	Source credential for gated or private Hugging Face repos.

Remote uploads retrieve

Get a remote upload job by ID. Poll the job until its status is REMOTE_UPLOAD_STATUS_SUCCEEDED.

Shell

tg models remote-uploads retrieve job_abc123

Alias: tg models remote-uploads get.

Parameters

Flag	Description
ID	(required) The remote upload job ID.

Remote uploads list

List the remote upload jobs in the current project.

Shell

tg models remote-uploads list

Alias: tg models remote-uploads ls.

Parameters

Flag	Description
--limit [number]	Maximum number of jobs to return.
--after [string]	Pagination cursor to start from.

Global options

Every command also accepts the global parameters, including --json for machine-readable output and --project to override the target project.

Resource IDs

Prefix	Resource
ml_	Model or adapter record.
cr_	Config revision (a deployable configuration of a model).
job_	Remote upload job.