diff --git a/aider/repomap.py b/aider/repomap.py
index 5c40c469b..e0827d71d 100644
--- a/aider/repomap.py
+++ b/aider/repomap.py
@@ -7,6 +7,7 @@ import sqlite3
 import sys
 import time
 import warnings
+import hashlib
 from collections import Counter, defaultdict, namedtuple
 from importlib import resources
 from pathlib import Path
@@ -223,6 +224,10 @@ class RepoMap:
     def save_tags_cache(self):
         pass
 
+    def _calculate_md5(self, content_bytes):
+        """Calculates the MD5 hash of byte content."""
+        return hashlib.md5(content_bytes).hexdigest()
+
     def get_mtime(self, fname):
         try:
             return os.path.getmtime(fname)
@@ -230,39 +235,106 @@ class RepoMap:
             self.io.tool_warning(f"File not found error: {fname}")
 
     def get_tags(self, fname, rel_fname):
-        # Check if the file is in the cache and if the modification time has not changed
-        file_mtime = self.get_mtime(fname)
-        if file_mtime is None:
-            return []
-
-        cache_key = fname
+        # 1. Get mtime and read content (handle errors)
         try:
-            val = self.TAGS_CACHE.get(cache_key)  # Issue #1308
-        except SQLITE_ERRORS as e:
-            self.tags_cache_error(e)
-            val = self.TAGS_CACHE.get(cache_key)
-
-        if val is not None and val.get("mtime") == file_mtime:
+            file_mtime = os.path.getmtime(fname)
+            # Read as bytes for consistent hashing
+            content_bytes = Path(fname).read_bytes()
+            # Decode for parsing, handle potential errors
+            code = content_bytes.decode(self.io.encoding, errors='replace')
+        except (FileNotFoundError, IsADirectoryError, OSError) as e:
+            # File inaccessible, warn and ensure it's removed from cache if present
+            if fname not in self.warned_files: # Avoid repeated warnings
+                self.io.tool_warning(f"RepoMap: Error accessing file {fname}: {e}")
+                self.warned_files.add(fname)
+            cache_key = fname
             try:
-                return self.TAGS_CACHE[cache_key]["data"]
-            except SQLITE_ERRORS as e:
-                self.tags_cache_error(e)
-                return self.TAGS_CACHE[cache_key]["data"]
+                # Use TAGS_CACHE.pop() which handles key errors gracefully
+                self.TAGS_CACHE.pop(cache_key, None)
+            except SQLITE_ERRORS as sql_e:
+                 self.tags_cache_error(sql_e)
+                 # If fallback to dict, try pop again
+                 if isinstance(self.TAGS_CACHE, dict):
+                     self.TAGS_CACHE.pop(cache_key, None)
+            return [] # Return empty list if file can't be accessed
 
-        # miss!
-        data = list(self.get_tags_raw(fname, rel_fname))
+        # 2. Calculate MD5
+        current_md5 = self._calculate_md5(content_bytes)
 
-        # Update the cache
+        # 3. Check cache
+        cache_key = fname
+        cached_data = None
         try:
-            self.TAGS_CACHE[cache_key] = {"mtime": file_mtime, "data": data}
-            self.save_tags_cache()
+            # Use get() with a default to avoid KeyError if key missing
+            cached_data = self.TAGS_CACHE.get(cache_key, None)
         except SQLITE_ERRORS as e:
             self.tags_cache_error(e)
-            self.TAGS_CACHE[cache_key] = {"mtime": file_mtime, "data": data}
+            # Try get() again after potential recovery
+            if isinstance(self.TAGS_CACHE, dict):
+                 cached_data = self.TAGS_CACHE.get(cache_key, None)
 
-        return data
+        # 4. Compare mtime and MD5
+        cache_valid = False
+        log_message = None
+        if cached_data:
+            cached_mtime = cached_data.get("mtime")
+            cached_md5 = cached_data.get("md5")
 
-    def get_tags_raw(self, fname, rel_fname):
+            if cached_mtime == file_mtime:
+                if cached_md5 == current_md5:
+                    # mtime and MD5 match, cache is valid
+                    cache_valid = True
+                else:
+                    # MD5 mismatch! Content changed despite same mtime.
+                    log_message = (
+                        f"RepoMap: Content change detected for {rel_fname} (MD5 mismatch,"
+                        f" mtime {file_mtime}). Re-parsing."
+                    )
+            else:
+                 # mtime mismatch - file definitely changed or cache is stale
+                 log_message = (
+                     f"RepoMap: File change detected for {rel_fname} (mtime mismatch: cached"
+                     f" {cached_mtime}, current {file_mtime}). Re-parsing."
+                 )
+
+        # 5. Return cached data or re-parse
+        if cache_valid:
+            try:
+                # Ensure data exists in the valid cache entry
+                return cached_data.get("data", [])
+            except Exception as e:
+                 # Handle potential issues reading cached data, force re-parse
+                 self.io.tool_warning(f"RepoMap: Error reading cached data for {fname}: {e}")
+                 cache_valid = False # Force re-parse
+
+        # Cache is invalid or file changed - log if needed and re-parse
+        if log_message:
+             self.io.tool_warning(log_message)
+
+        # Call the raw tag parsing function (passing the already read code)
+        tags_data = list(self.get_tags_raw(fname, rel_fname, code)) # Pass code here
+
+        # Update the cache with new mtime, md5, and data
+        try:
+            self.TAGS_CACHE[cache_key] = {
+                "mtime": file_mtime,
+                "md5": current_md5,
+                "data": tags_data,
+            }
+            # self.save_tags_cache() # save_tags_cache might be a no-op now with diskcache
+        except SQLITE_ERRORS as e:
+            self.tags_cache_error(e)
+            # Try saving to in-memory dict if disk cache failed
+            if isinstance(self.TAGS_CACHE, dict):
+                 self.TAGS_CACHE[cache_key] = {
+                     "mtime": file_mtime,
+                     "md5": current_md5,
+                     "data": tags_data,
+                 }
+
+        return tags_data
+
+    def get_tags_raw(self, fname, rel_fname, code):
         lang = filename_to_lang(fname)
         if not lang:
             return
@@ -271,22 +343,40 @@ class RepoMap:
             language = get_language(lang)
             parser = get_parser(lang)
         except Exception as err:
-            print(f"Skipping file {fname}: {err}")
+            # Use io object for output
+            self.io.tool_warning(f"Skipping file {fname} for tags: {err}")
             return
 
-        query_scm = get_scm_fname(lang)
-        if not query_scm.exists():
+        query_scm_path = get_scm_fname(lang) # Renamed variable for clarity
+        if not query_scm_path or not query_scm_path.exists():
+             # self.io.tool_warning(f"No tags query file found for language: {lang}") # Optional: more verbose logging
+             return
+        try:
+            query_scm = query_scm_path.read_text()
+        except Exception as e:
+            self.io.tool_warning(f"Error reading tags query file {query_scm_path}: {e}")
             return
-        query_scm = query_scm.read_text()
 
-        code = self.io.read_text(fname)
-        if not code:
+
+        # code = self.io.read_text(fname) # <-- REMOVE THIS LINE
+        if not code: # Check the passed code content
             return
-        tree = parser.parse(bytes(code, "utf-8"))
+        try:
+            tree = parser.parse(bytes(code, "utf-8"))
+        except Exception as e:
+            self.io.tool_warning(f"Error parsing {fname} with tree-sitter: {e}")
+            return
+
 
         # Run the tags queries
-        query = language.query(query_scm)
-        captures = query.captures(tree.root_node)
+        try:
+            query = language.query(query_scm)
+            captures = query.captures(tree.root_node)
+        except Exception as e:
+            # Catch errors during query execution as well
+            self.io.tool_warning(f"Error running tags query on {fname}: {e}")
+            return
+
 
         saw = set()
         if USING_TSL_PACK:
diff --git a/aider/resources/model-metadata.json b/aider/resources/model-metadata.json
index 72c394730..1b88463a2 100644
--- a/aider/resources/model-metadata.json
+++ b/aider/resources/model-metadata.json
@@ -204,6 +204,42 @@
         "supports_tool_choice": true,
         "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing"
     },
+    "gemini/gemini-2.5-pro-preview-05-06": {
+        "max_tokens": 8192,
+        "max_input_tokens": 1048576,
+        "max_output_tokens": 64000,
+        "max_images_per_prompt": 3000,
+        "max_videos_per_prompt": 10,
+        "max_video_length": 1,
+        "max_audio_length_hours": 8.4,
+        "max_audio_per_prompt": 1,
+        "max_pdf_size_mb": 30,
+        "input_cost_per_image": 0,
+        "input_cost_per_video_per_second": 0,
+        "input_cost_per_audio_per_second": 0,
+        "input_cost_per_token": 0.00000125,
+        "input_cost_per_character": 0,
+        "input_cost_per_token_above_128k_tokens": 0,
+        "input_cost_per_character_above_128k_tokens": 0,
+        "input_cost_per_image_above_128k_tokens": 0,
+        "input_cost_per_video_per_second_above_128k_tokens": 0,
+        "input_cost_per_audio_per_second_above_128k_tokens": 0,
+        "output_cost_per_token": 0.000010,
+        "output_cost_per_character": 0,
+        "output_cost_per_token_above_128k_tokens": 0,
+        "output_cost_per_character_above_128k_tokens": 0,
+        "litellm_provider": "gemini",
+        "mode": "chat",
+        "supports_system_messages": true,
+        "supports_function_calling": true,
+        "supports_vision": true,
+        "supports_audio_input": true,
+        "supports_video_input": true,
+        "supports_pdf_input": true,
+        "supports_response_schema": true,
+        "supports_tool_choice": true,
+        "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing"
+    },
     "vertex_ai/gemini-2.5-pro-exp-03-25": {
         "max_tokens": 8192,
         "max_input_tokens": 1048576,
diff --git a/aider/resources/model-settings.yml b/aider/resources/model-settings.yml
index 5b00a0f0d..f782f7059 100644
--- a/aider/resources/model-settings.yml
+++ b/aider/resources/model-settings.yml
@@ -963,6 +963,12 @@
   use_repo_map: true
   weak_model_name: gemini/gemini-2.0-flash
 
+- name: gemini/gemini-2.5-pro-preview-05-06
+  overeager: true
+  edit_format: diff-fenced
+  use_repo_map: true
+  weak_model_name: gemini/gemini-2.0-flash
+
 - name: gemini/gemini-2.5-pro-exp-03-25
   edit_format: diff-fenced
   use_repo_map: true
diff --git a/tests/basic/test_repomap.py b/tests/basic/test_repomap.py
index 322bac954..8dac835a5 100644
--- a/tests/basic/test_repomap.py
+++ b/tests/basic/test_repomap.py
@@ -4,6 +4,7 @@ import re
 import time
 import unittest
 from pathlib import Path
+from unittest.mock import patch
 
 import git
 
@@ -18,7 +19,167 @@ class TestRepoMap(unittest.TestCase):
     def setUp(self):
         self.GPT35 = Model("gpt-3.5-turbo")
 
+    # Helper function to calculate MD5 hash of a file
+    def _calculate_md5_for_file(self, file_path):
+        import hashlib
+        hasher = hashlib.md5()
+        with open(file_path, 'rb') as f:
+            while True:
+                chunk = f.read(8192)
+                if not chunk:
+                    break
+                hasher.update(chunk)
+        return hasher.hexdigest()
+
     def test_get_repo_map(self):
+        pass
+
+    @patch("aider.io.InputOutput.tool_warning")
+    def test_get_tags_md5_change_same_mtime(self, mock_tool_warning):
+        """Verify MD5 detection when mtime is unchanged."""
+        with GitTemporaryDirectory() as temp_dir:
+            # Create a test file
+            test_file = Path(temp_dir) / "test.py"
+            initial_content = "def func_a(): pass\n"
+            test_file.write_text(initial_content)
+            abs_path = str(test_file.resolve())
+            rel_path = "test.py"
+
+            # Initialize RepoMap and populate cache
+            io = InputOutput()
+            repo_map = RepoMap(main_model=self.GPT35, root=temp_dir, io=io)
+            initial_tags = repo_map.get_tags(abs_path, rel_path)
+            self.assertTrue(any(tag.name == "func_a" for tag in initial_tags))
+            initial_mtime = os.path.getmtime(abs_path)
+
+            # Modify content, reset mtime
+            new_content = "def func_b(): pass\n"
+            test_file.write_text(new_content)
+            os.utime(abs_path, (initial_mtime, initial_mtime)) # Reset mtime
+
+            # Call get_tags again
+            new_tags = repo_map.get_tags(abs_path, rel_path)
+
+            # Assertions
+            mock_tool_warning.assert_called_once()
+            self.assertIn("MD5 mismatch", mock_tool_warning.call_args[0][0])
+            self.assertTrue(any(tag.name == "func_b" for tag in new_tags))
+            self.assertFalse(any(tag.name == "func_a" for tag in new_tags))
+
+            # Check cache update
+            cached_data = repo_map.TAGS_CACHE.get(abs_path)
+            self.assertIsNotNone(cached_data)
+            expected_md5 = self._calculate_md5_for_file(abs_path)
+            self.assertEqual(cached_data.get("md5"), expected_md5)
+            self.assertEqual(cached_data.get("mtime"), initial_mtime)
+
+            del repo_map # Close cache
+
+    @patch("aider.io.InputOutput.tool_warning")
+    @patch("aider.repomap.RepoMap.get_tags_raw")
+    def test_get_tags_no_change(self, mock_get_tags_raw, mock_tool_warning):
+        """Verify cache is used when file is unchanged."""
+        with GitTemporaryDirectory() as temp_dir:
+            test_file = Path(temp_dir) / "test.py"
+            initial_content = "def func_a(): pass\n"
+            test_file.write_text(initial_content)
+            abs_path = str(test_file.resolve())
+            rel_path = "test.py"
+
+            io = InputOutput()
+            repo_map = RepoMap(main_model=self.GPT35, root=temp_dir, io=io)
+
+            # Initial call to populate cache
+            initial_tags = repo_map.get_tags(abs_path, rel_path)
+            mock_get_tags_raw.assert_called_once() # Called once initially
+            mock_get_tags_raw.reset_mock() # Reset for the next check
+
+            # Call get_tags again without changes
+            second_tags = repo_map.get_tags(abs_path, rel_path)
+
+            # Assertions
+            mock_tool_warning.assert_not_called()
+            mock_get_tags_raw.assert_not_called() # Should not be called again
+            self.assertEqual(initial_tags, second_tags)
+
+            del repo_map # Close cache
+
+    @patch("aider.io.InputOutput.tool_warning")
+    def test_get_tags_mtime_change(self, mock_tool_warning):
+        """Verify standard mtime-based change detection still works."""
+        with GitTemporaryDirectory() as temp_dir:
+            test_file = Path(temp_dir) / "test.py"
+            initial_content = "def func_a(): pass\n"
+            test_file.write_text(initial_content)
+            abs_path = str(test_file.resolve())
+            rel_path = "test.py"
+
+            io = InputOutput()
+            repo_map = RepoMap(main_model=self.GPT35, root=temp_dir, io=io)
+
+            # Initial call
+            initial_tags = repo_map.get_tags(abs_path, rel_path)
+            self.assertTrue(any(tag.name == "func_a" for tag in initial_tags))
+
+            # Modify content (mtime will change naturally)
+            time.sleep(0.01) # Ensure mtime is different
+            new_content = "def func_b(): pass\n"
+            test_file.write_text(new_content)
+            new_mtime = os.path.getmtime(abs_path)
+
+            # Call get_tags again
+            new_tags = repo_map.get_tags(abs_path, rel_path)
+
+            # Assertions
+            mock_tool_warning.assert_called_once()
+            self.assertIn("mtime mismatch", mock_tool_warning.call_args[0][0])
+            self.assertTrue(any(tag.name == "func_b" for tag in new_tags))
+            self.assertFalse(any(tag.name == "func_a" for tag in new_tags))
+
+            # Check cache update
+            cached_data = repo_map.TAGS_CACHE.get(abs_path)
+            self.assertIsNotNone(cached_data)
+            expected_md5 = self._calculate_md5_for_file(abs_path)
+            self.assertEqual(cached_data.get("md5"), expected_md5)
+            self.assertEqual(cached_data.get("mtime"), new_mtime)
+
+            del repo_map # Close cache
+
+    @patch("aider.io.InputOutput.tool_warning")
+    def test_get_tags_file_not_found_after_cache(self, mock_tool_warning):
+        """Verify graceful handling if a cached file becomes inaccessible."""
+        with GitTemporaryDirectory() as temp_dir:
+            test_file = Path(temp_dir) / "test.py"
+            test_file.write_text("def func_a(): pass\n")
+            abs_path = str(test_file.resolve())
+            rel_path = "test.py"
+
+            io = InputOutput()
+            repo_map = RepoMap(main_model=self.GPT35, root=temp_dir, io=io)
+
+            # Populate cache
+            repo_map.get_tags(abs_path, rel_path)
+            self.assertIn(abs_path, repo_map.TAGS_CACHE)
+
+            # Delete the file
+            os.remove(abs_path)
+
+            # Call get_tags again
+            result = repo_map.get_tags(abs_path, rel_path)
+
+            # Assertions
+            mock_tool_warning.assert_called()
+            # Check if any call contains "Error accessing file" or "FileNotFoundError"
+            warning_found = any(
+                "Error accessing file" in call[0][0] or "FileNotFoundError" in call[0][0]
+                for call in mock_tool_warning.call_args_list
+            )
+            self.assertTrue(warning_found, "Expected file access error warning not found")
+
+            self.assertEqual(result, [])
+            self.assertNotIn(abs_path, repo_map.TAGS_CACHE)
+
+            del repo_map # Close cache
         # Create a temporary directory with sample files for testing
         test_files = [
             "test_file1.py",