Optionalapply_Whether to apply library patches for improved compatibility.
Optionalbase_Use a custom Docker base image instead of the default Truss image.
OptionalbdnOptionalbis_Configuration options for BIS LLM deployments. This field may change in the future.
OptionalbuildOptionalbuild_A list of shell commands to run during Docker build. These commands execute after system packages and Python requirements are installed.
Optionalbundled_The folder for custom packages in your Truss.
Optionalcache_Optionaldata_The folder for data files in your Truss.
OptionaldescriptionA description of your model.
Optionaldocker_Deploy a custom Docker image that has its own HTTP server, without writing a Model class.
Optionalenvironment_Key-value pairs exposed to the environment that the model executes in. Do not store secret values here.
Optionalexamples_Path to a file containing example model inputs.
Optionalexternal_External data to be downloaded and made available under the data directory at serving time.
Optionalexternal_Use external_package_dirs to access custom packages located outside your Truss. This lets multiple Trusses share the same package.
Optionalinput_Optionallive_If true, changes to your model code are automatically reloaded without restarting the server.
Optionalmodel_Optionalmodel_Optionalmodel_The name of the class that defines your Truss model. This class must implement at least a predict method.
Optionalmodel_Optionalmodel_A flexible field for additional metadata. The entire config file is available to your model at runtime.
Optionalexample_model_input?: { [k: string]: unknown }Sample input that populates the Baseten playground.
Optionalmodel_The folder containing your model class.
Optionalmodel_The name of your model. This is displayed in the model details page in the Baseten UI.
Optionalmodel_Optionalpython_The Python version to use.
OptionalrequirementsA list of Python dependencies in pip requirements file format. Mutually exclusive with 'requirements_file'.
Optionalrequirements_Path to a dependency file. Supports requirements.txt, pyproject.toml, and uv.lock. Mutually exclusive with 'requirements'.
OptionalresourcesOptionalruntimeOptionalsecretsDeclare secrets your model needs at runtime, such as API keys or access tokens. Use null as a placeholder; store actual values in your organization settings.
Optionalspec_Optionalsystem_System packages that you would typically install using apt on a Debian operating system.
Optionaltraining_Configuration for deploying from training checkpoints.
Optionaltrt_TensorRT-LLM configuration for optimized LLM inference.
Optionaluse_Optionalweights
Configuration for a Truss model deployment.