diff --git a/refact-server/refact_webgui/webgui/selfhost_fastapi_completions.py b/refact-server/refact_webgui/webgui/selfhost_fastapi_completions.py
index 09ad460b1..b76c64abd 100644
--- a/refact-server/refact_webgui/webgui/selfhost_fastapi_completions.py
+++ b/refact-server/refact_webgui/webgui/selfhost_fastapi_completions.py
@@ -2,7 +2,6 @@
 import json
 import copy
 import asyncio
-import aiohttp
 import aiofiles
 import termcolor
 import os
@@ -13,6 +12,8 @@
 from fastapi import APIRouter, HTTPException, Query, Header
 from fastapi.responses import Response, StreamingResponse
 
+from itertools import chain
+
 from refact_utils.scripts import env
 from refact_utils.finetune.utils import running_models_and_loras
 from refact_utils.third_party.utils.models import available_third_party_models
@@ -250,8 +251,9 @@ def _select_default_model(models: List[str]) -> str:
         # completion models
         completion_models = {}
         for model_name in running_models.get("completion", []):
-            if model_info := self._model_assigner.models_db.get(_get_base_model_info(model_name)):
-                completion_models[model_name] = self._model_assigner.to_completion_model_record(model_info)
+            base_model_name = _get_base_model_info(model_name)
+            if model_info := self._model_assigner.models_db.get(base_model_name):
+                completion_models[model_name] = self._model_assigner.to_completion_model_record(base_model_name, model_info)
             elif model := available_third_party_models().get(model_name):
                 completion_models[model_name] = model.to_completion_model_record()
             else:
@@ -261,8 +263,9 @@ def _select_default_model(models: List[str]) -> str:
         # chat models
         chat_models = {}
         for model_name in running_models.get("chat", []):
-            if model_info := self._model_assigner.models_db.get(_get_base_model_info(model_name)):
-                chat_models[model_name] = self._model_assigner.to_chat_model_record(model_info)
+            base_model_name = _get_base_model_info(model_name)
+            if model_info := self._model_assigner.models_db.get(base_model_name):
+                chat_models[model_name] = self._model_assigner.to_chat_model_record(base_model_name, model_info)
             elif model := available_third_party_models().get(model_name):
                 chat_models[model_name] = model.to_chat_model_record()
             else:
@@ -337,6 +340,18 @@ def _parse_client_version(user_agent: str = Header(None)) -> Optional[Tuple[int,
 
     @staticmethod
     def _to_deprecated_caps_format(data: Dict[str, Any]):
+        models_dict_patch = {}
+        for model_name, model_record in chain(
+                data["completion"]["models"].items(),
+                data["completion"]["models"].items(),
+        ):
+            dict_patch = {}
+            if n_ctx := model_record.get("n_ctx"):
+                dict_patch["n_ctx"] = n_ctx
+            if supports_tools := model_record.get("supports_tools"):
+                dict_patch["supports_tools"] = supports_tools
+            if dict_patch:
+                models_dict_patch[model_name] = dict_patch
         return {
             "cloud_name": data["cloud_name"],
             "endpoint_template": data["completion"]["endpoint"],
@@ -349,7 +364,7 @@ def _to_deprecated_caps_format(data: Dict[str, Any]):
             "code_completion_default_model": data["completion"]["default_model"],
             "multiline_code_completion_default_model": data["completion"]["default_multiline_model"],
             "code_chat_default_model": data["chat"]["default_model"],
-            "models_dict_patch": {},  # NOTE: this actually should have n_ctx, but we're skiping it
+            "models_dict_patch": models_dict_patch,
             "default_embeddings_model": data["embedding"]["default_model"],
             "endpoint_embeddings_template": "v1/embeddings",
             "endpoint_embeddings_style": "openai",
diff --git a/refact-server/refact_webgui/webgui/selfhost_model_assigner.py b/refact-server/refact_webgui/webgui/selfhost_model_assigner.py
index 69024743a..192de301f 100644
--- a/refact-server/refact_webgui/webgui/selfhost_model_assigner.py
+++ b/refact-server/refact_webgui/webgui/selfhost_model_assigner.py
@@ -107,17 +107,15 @@ def share_gpu_backends(self) -> Set[str]:
     def models_db(self) -> Dict[str, Any]:
         return models_mini_db
 
-    @staticmethod
-    def to_completion_model_record(model_info: Dict[str, Any]) -> Dict[str, Any]:
+    def to_completion_model_record(self, model_name: str, model_info: Dict[str, Any]) -> Dict[str, Any]:
         return {
-            "n_ctx": model_info["T"],
+            "n_ctx": min(self.model_assignment["model_assign"].get(model_name, {}).get("n_ctx", model_info["T"]), model_info["T"]),
             "supports_scratchpads": model_info["supports_scratchpads"]["completion"],
         }
 
-    @staticmethod
-    def to_chat_model_record(model_info: Dict[str, Any]) -> Dict[str, Any]:
+    def to_chat_model_record(self, model_name: str, model_info: Dict[str, Any]) -> Dict[str, Any]:
         return {
-            "n_ctx": model_info["T"],
+            "n_ctx": min(self.model_assignment["model_assign"].get(model_name, {}).get("n_ctx", model_info["T"]), model_info["T"]),
             "supports_scratchpads": model_info["supports_scratchpads"]["chat"],
         }