diff --git a/benchmarks/cli.py b/benchmarks/cli.py
index 4e6cab1..26962de 100644
--- a/benchmarks/cli.py
+++ b/benchmarks/cli.py
@@ -26,6 +26,16 @@ def cmd_run(experiments: Dict[str, ExperimentBuilder[Experiment]], args):
     experiments[args.experiment].build().run()
 
 
+def cmd_describe(args):
+    if not args.type:
+        print(f'Available experiment types are:')
+        for experiment in config_parser.experiment_types.keys():
+            print(f'  - {experiment}')
+        return
+
+    print(config_parser.experiment_types[args.type].schema_json(indent=2))
+
+
 def _parse_config(config: Path) -> Dict[str, ExperimentBuilder[Experiment]]:
     if not config.exists():
         print(f'Config file {config} does not exist.')
@@ -52,21 +62,31 @@ def _init_logging():
 
 def main():
     parser = argparse.ArgumentParser()
-    parser.add_argument('config', type=Path, help="Path to the experiment configuration file.")
 
     commands = parser.add_subparsers(required=True)
-    list_cmd = commands.add_parser('list', help='Lists available experiments.')
-    list_cmd.set_defaults(func=cmd_list)
 
-    run_cmd = commands.add_parser('run')
+    experiments = commands.add_parser('experiments', help='List or run experiments in config file.')
+    experiments.add_argument('config', type=Path, help='Path to the experiment configuration file.')
+    experiment_commands = experiments.add_subparsers(required=True)
+
+    list_cmd = experiment_commands.add_parser('list', help='Lists available experiments.')
+    list_cmd.set_defaults(func=lambda args: cmd_list(_parse_config(args.config), args))
+
+    run_cmd = experiment_commands.add_parser('run', help='Runs an experiment')
     run_cmd.add_argument('experiment', type=str, help='Name of the experiment to run.')
-    run_cmd.set_defaults(func=cmd_run)
+    run_cmd.set_defaults(func=lambda args: cmd_run(_parse_config(args.config), args))
+
+    describe = commands.add_parser('describe', help='Shows the JSON schema for the various experiment types.')
+    describe.add_argument('type', type=str, help='Type of the experiment to describe.',
+                          choices=config_parser.experiment_types.keys(), nargs='?')
+
+    describe.set_defaults(func=cmd_describe)
 
     args = parser.parse_args()
 
     _init_logging()
 
-    args.func(_parse_config(args.config), args)
+    args.func(args)
 
 
 if __name__ == '__main__':
diff --git a/benchmarks/core/config.py b/benchmarks/core/config.py
index c0da879..d88237f 100644
--- a/benchmarks/core/config.py
+++ b/benchmarks/core/config.py
@@ -57,13 +57,13 @@ class ConfigParser:
     """
 
     def __init__(self):
-        self.root_tags = {}
+        self.experiment_types = {}
 
     def register(self, root: Type[ExperimentBuilder[TExperiment]]):
         name = root.__name__
         alias = cast(Callable[[str], str],
                      root.model_config.get('alias_generator', lambda x: x))(name)
-        self.root_tags[alias] = root
+        self.experiment_types[alias] = root
 
     @overload
     def parse(self, data: dict) -> Dict[str, ExperimentBuilder[TExperiment]]:
@@ -80,6 +80,6 @@ class ConfigParser:
             entries = data
 
         return {
-            tag: self.root_tags[tag].model_validate(config)
+            tag: self.experiment_types[tag].model_validate(config)
             for tag, config in entries.items()
         }