diff --git a/src/raman_fitting/config/default_models/first_order.toml b/src/raman_fitting/config/default_models/first_order.toml
index 943cd05..321d650 100644
--- a/src/raman_fitting/config/default_models/first_order.toml
+++ b/src/raman_fitting/config/default_models/first_order.toml
@@ -1,4 +1,3 @@
-
 [first_order]
 
 [first_order.models]
diff --git a/src/raman_fitting/config/default_models/second_order.toml b/src/raman_fitting/config/default_models/second_order.toml
index 34113e3..1697d02 100644
--- a/src/raman_fitting/config/default_models/second_order.toml
+++ b/src/raman_fitting/config/default_models/second_order.toml
@@ -1,4 +1,3 @@
-
 [second_order]
 
 [second_order.models]
diff --git a/src/raman_fitting/config/default_models/spectrum_windows.toml b/src/raman_fitting/config/default_models/spectrum_windows.toml
new file mode 100644
index 0000000..f58986f
--- /dev/null
+++ b/src/raman_fitting/config/default_models/spectrum_windows.toml
@@ -0,0 +1,10 @@
+[spectrum]
+
+[spectrum.windows]
+full = {"min" = 200, "max" = 3600}
+full_first_and_second = {"min" = 800, "max" = 3500}
+low = {"min" = 150, "max" = 850, "extra_margin" = 10}
+first_order = {"min" = 900, "max" = 2000}
+mid = {"min" = 1850, "max" = 2150, "extra_margin" = 10}
+normalization = {"min" = 1500, "max" = 1675, "extra_margin" = 10}
+second_order = {"min" = 2150, "max" = 3380}
\ No newline at end of file
diff --git a/src/raman_fitting/models/deconvolution/spectrum_regions.py b/src/raman_fitting/models/deconvolution/spectrum_regions.py
new file mode 100644
index 0000000..c995ba8
--- /dev/null
+++ b/src/raman_fitting/models/deconvolution/spectrum_regions.py
@@ -0,0 +1,20 @@
+from enum import StrEnum
+from typing import Dict
+
+from pydantic import BaseModel
+from raman_fitting.config.default_models import load_config_from_toml_files
+
+
+def get_default_regions_from_toml_files() -> Dict[str, Dict[str, float]]:
+    default_windows = load_config_from_toml_files().get("spectrum", {}).get("windows", {})
+    return default_windows
+
+
+WindowNames = StrEnum('WindowNames', " ".join(get_default_regions_from_toml_files()), module=__name__)
+
+
+class SpectrumWindowLimits(BaseModel):
+    name: WindowNames
+    min: int
+    max: int
+    extra_margin: int = 20
diff --git a/src/raman_fitting/models/fit_models.py b/src/raman_fitting/models/fit_models.py
index 75028b6..fda4540 100644
--- a/src/raman_fitting/models/fit_models.py
+++ b/src/raman_fitting/models/fit_models.py
@@ -7,6 +7,8 @@
 from lmfit.model import ModelResult
 
 from raman_fitting.models.deconvolution.base_model import BaseLMFitModel
+from raman_fitting.models.deconvolution.spectrum_regions import WindowNames
+
 from raman_fitting.models.spectrum import SpectrumData
 
 logger = logging.getLogger(__name__)
@@ -29,6 +31,7 @@ class SpectrumFitModel(BaseModel):
 
     spectrum: SpectrumData
     model: BaseLMFitModel
+    window: WindowNames
     fit_kwargs: Dict = Field(default_factory=dict)
     fit_result: ModelResult = Field(None, init_var=False)
     elapsed_time: float = Field(0, init_var=False, repr=False)
@@ -43,7 +46,7 @@ def match_window_names(self):
             )
         return self
 
-    def run_fit(self) -> ModelResult:
+    def run_fit(self) -> None:
         if "method" not in self.fit_kwargs:
             self.fit_kwargs["method"] = "leastsq"
         lmfit_model = self.model.lmfit_model
@@ -54,6 +57,7 @@ def run_fit(self) -> ModelResult:
         self.elapsed_time = elapsed_seconds
         self.fit_result = fit_result
 
+
     def process_fit_results(self):
         #  TODO add parameter post processing steps
         self.fit_result
@@ -77,7 +81,7 @@ def run_fit(
 
 
 if __name__ == "__main__":
-    from raman_fitting.config.settings import settings
+    from raman_fitting.config.base_settings import settings
 
     test_fixtures = list(settings.internal_paths.example_fixtures.glob("*txt"))
     file = [i for i in test_fixtures if "_pos4" in i.stem][0]
diff --git a/src/raman_fitting/models/splitter.py b/src/raman_fitting/models/splitter.py
index ab10330..afa4683 100644
--- a/src/raman_fitting/models/splitter.py
+++ b/src/raman_fitting/models/splitter.py
@@ -1,38 +1,11 @@
-from enum import Enum
 from typing import Dict, Any
 import numpy as np
 
 from pydantic import BaseModel, model_validator, Field
 from .spectrum import SpectrumData
+from .deconvolution.spectrum_regions import SpectrumWindowLimits, WindowNames, get_default_regions_from_toml_files
 
 
-class WindowNames(str, Enum):
-    full = "full"
-    full_first_and_second = "full_first_and_second"
-    low = "low"
-    first_order = "first_order"
-    mid = "mid"
-    second_order = "second_order"
-    normalization = "normalization"
-
-
-SPECTRUM_WINDOWS_LIMITS = {
-    WindowNames.full: {"min": 200, "max": 3600},
-    WindowNames.full_first_and_second: {"min": 800, "max": 3500},
-    WindowNames.low: {"min": 150, "max": 850, "extra_margin": 10},
-    WindowNames.first_order: {"min": 900, "max": 2000},
-    WindowNames.mid: {"min": 1850, "max": 2150, "extra_margin": 10},
-    WindowNames.second_order: {"min": 2150, "max": 3380},
-    WindowNames.normalization: {"min": 1500, "max": 1675, "extra_margin": 10},
-}
-
-
-class SpectrumWindowLimits(BaseModel):
-    name: WindowNames
-    min: int
-    max: int
-    extra_margin: int = 20
-
 
 class SplittedSpectrum(BaseModel):
     spectrum: SpectrumData
@@ -68,10 +41,11 @@ def get_window(self, window_name: WindowNames):
         return self.spec_windows[spec_window_key]
 
 
-def get_default_spectrum_window_limits() -> Dict[str, SpectrumWindowLimits]:
+def get_default_spectrum_window_limits(windows_mapping: Dict = None) -> Dict[str, SpectrumWindowLimits]:
+    if windows_mapping is None:
+        windows_mapping = get_default_regions_from_toml_files()
     windows = {}
-    for window_type, window_config in SPECTRUM_WINDOWS_LIMITS.items():
-        window_name = window_type.name
+    for window_name, window_config in windows_mapping.items():
         windows[window_name] = SpectrumWindowLimits(name=window_name, **window_config)
     return windows