diff --git a/src/eynollah/cli/cli_layout.py b/src/eynollah/cli/cli_layout.py
index 8ce0872..03cf9c8 100644
--- a/src/eynollah/cli/cli_layout.py
+++ b/src/eynollah/cli/cli_layout.py
@@ -165,6 +165,11 @@ import click
     type=click.IntRange(min=0),
     help="number of parallel images to process (also helps better utilise GPU if available); 0 means based on autodetected number of processor cores",
 )
+@click.option(
+    "--device",
+    "-D",
+    help="placement of computations in predictors for each model type; if none (by default), will try to use first available GPU or fall back to CPU; set string to force using a device (e.g. 'GPU0', 'GPU1' or 'CPU'). Can also be a comma-separated list of model category to device mappings (e.g. 'col_classifier:CPU,page:GPU0,*:GPU1')",
+)
 @click.pass_context
 def layout_cli(
     ctx,
@@ -194,6 +199,7 @@ def layout_cli(
     skip_layout_and_reading_order,
     ignore_page_extraction,
     num_jobs,
+    device,
 ):
     """
     Detect Layout (with optional image enhancement and reading order detection)
@@ -209,6 +215,7 @@ def layout_cli(
     assert bool(image) != bool(dir_in), "Either -i (single input) or -di (directory) must be provided, but not both."
     eynollah = Eynollah(
         model_zoo=ctx.obj.model_zoo,
+        device=device,
         enable_plotting=enable_plotting,
         allow_enhancement=allow_enhancement,
         curved_line=curved_line,
diff --git a/src/eynollah/eynollah.py b/src/eynollah/eynollah.py
index 4823559..67ee525 100644
--- a/src/eynollah/eynollah.py
+++ b/src/eynollah/eynollah.py
@@ -121,6 +121,7 @@ class Eynollah:
         self,
         *,
         model_zoo: EynollahModelZoo,
+        device: str = '',
         enable_plotting : bool = False,
         allow_enhancement : bool = False,
         curved_line : bool = False,
@@ -165,10 +166,10 @@ class Eynollah:
         t_start = time.time()
 
         self.logger.info("Loading models...")
-        self.setup_models()
+        self.setup_models(device=device)
         self.logger.info(f"Model initialization complete ({time.time() - t_start:.1f}s)")
 
-    def setup_models(self):
+    def setup_models(self, device=''):
 
         # load models, depending on modes
         # (note: loading too many models can cause OOM on GPU/CUDA,
@@ -194,7 +195,7 @@ class Eynollah:
         if self.tables:
             loadable.append("table")
 
-        self.model_zoo.load_models(*loadable)
+        self.model_zoo.load_models(*loadable, device=device)
         for model in loadable:
             # retrieve and cache output shapes
             if model.endswith(('_resized', '_patched')):
diff --git a/src/eynollah/model_zoo/model_zoo.py b/src/eynollah/model_zoo/model_zoo.py
index 92fedfb..b72d36a 100644
--- a/src/eynollah/model_zoo/model_zoo.py
+++ b/src/eynollah/model_zoo/model_zoo.py
@@ -3,6 +3,7 @@ import json
 import logging
 from copy import deepcopy
 from pathlib import Path
+from fnmatch import fnmatchcase
 from typing import Dict, List, Optional, Tuple, Type, Union
 
 from tabulate import tabulate
@@ -74,6 +75,7 @@ class EynollahModelZoo:
     def load_models(
         self,
         *all_load_args: Union[str, Tuple[str], Tuple[str, str], Tuple[str, str, str]],
+        device: str = '',
     ) -> Dict:
         """
         Load all models by calling load_model and return a dictionary mapping model_category to loaded model
@@ -93,7 +95,7 @@ class EynollahModelZoo:
                 load_args[0] = model_category[:-8]
                 load_kwargs["patched"] = True
             ret[model_category] = Predictor(self.logger, self)
-            ret[model_category].load_model(*load_args, **load_kwargs)
+            ret[model_category].load_model(*load_args, **load_kwargs, device=device)
         self._loaded.update(ret)
         return self._loaded
 
@@ -104,6 +106,7 @@ class EynollahModelZoo:
         model_path_override: Optional[str] = None,
             patched: bool = False,
             resized: bool = False,
+            device: str = '',
     ) -> AnyModel:
         """
         Load any model
@@ -123,10 +126,24 @@ class EynollahModelZoo:
         )
         cuda = False
         try:
-            device = tf.config.list_physical_devices('GPU')[0]
-            tf.config.experimental.set_memory_growth(device, True)
-            cuda = True
-            self.logger.info("using GPU %s", device.name)
+            gpus = tf.config.list_physical_devices('GPU')
+            if device:
+                if ',' in device:
+                    for spec in device.split(','):
+                        cat, dev = spec.split(':')
+                        if fnmatchcase(model_category, cat):
+                            device = dev
+                            break
+                if device == 'CPU':
+                    gpus = []
+                else:
+                    assert device.startswith('GPU')
+                    gpus = [gpus[int(device[3:])]]
+            tf.config.set_visible_devices(gpus, 'GPU')
+            for device in gpus:
+                tf.config.experimental.set_memory_growth(device, True)
+                cuda = True
+                self.logger.info("using GPU %s for model %s", device.name, model_category)
         except RuntimeError:
             self.logger.exception("cannot configure GPU devices")
         if not cuda: