diff --git a/crawl4ai/__version__.py b/crawl4ai/__version__.py
index 9b49905e7..352aa28ab 100644
--- a/crawl4ai/__version__.py
+++ b/crawl4ai/__version__.py
@@ -1,7 +1,7 @@
 # crawl4ai/__version__.py
 
 # This is the version that will be used for stable releases
-__version__ = "0.8.6"
+__version__ = "0.8.7a5"
 
 # For nightly builds, this gets set during build process
 __nightly_version__ = None
diff --git a/crawl4ai/cli.py b/crawl4ai/cli.py
index 02b67155e..e20d22929 100644
--- a/crawl4ai/cli.py
+++ b/crawl4ai/cli.py
@@ -35,7 +35,7 @@
 from crawl4ai.browser_profiler import ShrinkLevel, _format_size
 from crawl4ai.config import USER_SETTINGS
 from crawl4ai.cloud import cloud_cmd
-from litellm import completion
+from nanollm import completion
 from pathlib import Path
 
 
@@ -66,7 +66,7 @@ def setup_llm_config() -> tuple[str, str]:
     if not provider:
         click.echo("\nNo default LLM provider configured.")
         click.echo("Provider format: 'company/model' (e.g., 'openai/gpt-4o', 'anthropic/claude-3-sonnet')")
-        click.echo("See available providers at: https://docs.litellm.ai/docs/providers")
+        click.echo("See available providers at: https://github.com/unclecode/nanollm#supported-providers")
         provider = click.prompt("Enter provider")
         
     if not provider.startswith("ollama/"):
@@ -344,7 +344,7 @@ def show_examples():
       - cohere/command
       - google/gemini-pro
     
-    See full list of providers: https://docs.litellm.ai/docs/providers
+    See full list of providers: https://github.com/unclecode/nanollm#supported-providers
     
     # Set default LLM provider and token in advance
     crwl config set DEFAULT_LLM_PROVIDER "anthropic/claude-3-sonnet"
diff --git a/crawl4ai/legacy/llmtxt.py b/crawl4ai/legacy/llmtxt.py
index 302564165..a06975b87 100644
--- a/crawl4ai/legacy/llmtxt.py
+++ b/crawl4ai/legacy/llmtxt.py
@@ -11,14 +11,14 @@
 from nltk.tokenize import word_tokenize
 from nltk.corpus import stopwords
 from nltk.stem import WordNetLemmatizer
-from litellm import batch_completion
+from nanollm import batch_completion
 from .async_logger import AsyncLogger
-import litellm
+import nanollm
 import pickle
 import hashlib  # <--- ADDED for file-hash
 import glob
 
-litellm.set_verbose = False
+nanollm.set_verbose = False
 
 
 def _compute_file_hash(file_path: Path) -> str:
diff --git a/crawl4ai/utils.py b/crawl4ai/utils.py
index 4b3d96906..962835b5d 100644
--- a/crawl4ai/utils.py
+++ b/crawl4ai/utils.py
@@ -1774,10 +1774,10 @@ def perform_completion_with_backoff(
         dict: The API response or an error message after all retries.
     """
 
-    from litellm import completion
-    from litellm.exceptions import RateLimitError
-    import litellm
-    litellm.drop_params = True  # Auto-drop unsupported params (e.g., temperature for O-series/GPT-5)
+    from nanollm import completion
+    from nanollm.exceptions import RateLimitError
+    import nanollm
+    nanollm.drop_params = True  # Auto-drop unsupported params (e.g., temperature for O-series/GPT-5)
 
     extra_args = {"temperature": 0.01, "api_key": api_token, "base_url": base_url}
     if json_response:
@@ -1866,11 +1866,11 @@ async def aperform_completion_with_backoff(
         dict: The API response or an error message after all retries.
     """
 
-    from litellm import acompletion
-    from litellm.exceptions import RateLimitError
-    import litellm
+    from nanollm import acompletion
+    from nanollm.exceptions import RateLimitError
+    import nanollm
     import asyncio
-    litellm.drop_params = True  # Auto-drop unsupported params (e.g., temperature for O-series/GPT-5)
+    nanollm.drop_params = True  # Auto-drop unsupported params (e.g., temperature for O-series/GPT-5)
 
     extra_args = {"temperature": 0.01, "api_key": api_token, "base_url": base_url}
     if json_response:
@@ -1991,7 +1991,7 @@ def extract_blocks_batch(batch_data, provider="groq/llama3-70b-8192", api_token=
     """
 
     api_token = os.getenv("GROQ_API_KEY", None) if not api_token else api_token
-    from litellm import batch_completion
+    from nanollm import batch_completion
 
     messages = []
 
@@ -3566,9 +3566,9 @@ async def get_text_embeddings(
     if not texts:
         return np.array([])
     
-    # If LLMConfig provided, use litellm for embeddings
+    # If LLMConfig provided, use nanollm for embeddings
     if llm_config is not None:
-        from litellm import aembedding
+        from nanollm import aembedding
         
         # Get embedding model from config or use default
         embedding_model = llm_config.get('provider', 'text-embedding-3-small')
diff --git a/pyproject.toml b/pyproject.toml
index ee237d5b3..6f1d04596 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -18,7 +18,7 @@ dependencies = [
     "aiosqlite~=0.20",
     "anyio>=4.0.0",
     "lxml~=5.3",
-    "unclecode-litellm==1.81.13",
+    "nanollm @ git+https://github.com/hafezparast/nanollm-final.git@8a8d93b02a77f04a32f202620d95c52ad26c7d05",
     "numpy>=1.26.0,<3",
     "pillow>=10.4",
     "playwright>=1.49.0",
@@ -46,7 +46,9 @@ dependencies = [
     "humanize>=4.10.0",
     "lark>=1.2.2",
     "alphashape>=1.3.1",
-    "shapely>=2.0.0"
+    "shapely>=2.0.0",
+    "packaging>=21.0",
+    "tiktoken>=0.5.0"
 ]
 classifiers = [
     "Development Status :: 4 - Beta",
diff --git a/requirements.txt b/requirements.txt
index 9686ffc95..17b1a18a2 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -5,7 +5,7 @@ aiohttp>=3.11.11
 aiosqlite~=0.20
 anyio>=4.0.0
 lxml~=5.3
-unclecode-litellm==1.81.13
+nanollm @ git+https://github.com/hafezparast/nanollm.git@v0.1.1
 numpy>=1.26.0,<3
 pillow>=10.4
 playwright>=1.49.0