https://api.synapsai.cloud/v1/model-artifacts. The key needs the Upload artifacts permission.
When the artifact status is ready, deploy it from the launch wizard by choosing that artifact as the model source. The pipeline you set must be one of the supported tasks. Safetensors weights are supported. Custom code is supported when Transformers, Diffusers, or vLLM can load it in the standard way.
Create and upload
1
Create the artifact record
Send
display_name and a pipeline such as text-generation. The response includes artifact.id.2
Start an upload session
POST /v1/model-artifacts/{artifact_id}/uploads returns upload_id. Another session can be started while an upload is already in progress; it adds files and does not replace what is already uploading. The artifact cannot be in use by a deployment.3
Send each file
PUT /v1/model-artifacts/{artifact_id}/uploads/{upload_id}/files?path= streams the raw bytes. path is the relative path stored in the artifact (config.json, model.safetensors).4
Complete the upload
POST .../complete stores the files and marks the artifact ready. Weight, pipeline, and config checks run when the artifact is deployed.Python SDK
pip install --upgrade synapsai-python. client.model_artifacts creates the record, uploads a file or a directory, and reads the artifact back.
upload() starts a session, sends every file, and completes it. Directory uploads keep relative paths. A single file is stored under its filename. pipeline is set only by create().
AsyncSynapsAI has these methods too. Raise timeout for large weight files. The example above uses 3600 seconds.
Step by step
CLI
The package installs asynapsai command. The artifact must already exist.
--timeout defaults to 3600 seconds.
API reference
A 409 means the artifact is attached to a deployment. A 403 on start means the storage quota is full. Ephemeral upload keys cannot create a new artifact record.