Add --parser CLI arg and litellm dependency for future integration

- Added litellm>=1.40.0 to requirements.txt
- Added --parser argument (auto/litellm, default auto)

Note: Full litellm integration requires significant refactoring of the
chat completion endpoints to use litellm.completion() for standardized
responses, adding rate limit headers, and error handling.

coderai

@@ -6287,6 +6287,13 @@ def parse_args():
         default=None,
         help="Path to store generated files (images, audio). If specified, files will be saved here and served over web.",
     )
+    parser.add_argument(
+        "--parser",
+        type=str,
+        default="auto",
+        choices=["auto", "litellm"],
+        help="Tool call parser to use: 'auto' for internal parser, 'litellm' for LiteLLM's parser. Default: auto",
+    )
     return parser.parse_args()
 def main():
     """Main entry point."""

requirements.txt

@@ -44,6 +44,9 @@ procname>=0.3.0
 faster-whisper>=0.10.0  # For NVIDIA/CUDA whisper transcription
 whispercpp>=1.0.0  # Alternative whisper library (works without PyTorch)
 
+# LiteLLM for standardized API responses
+litellm>=1.40.0
+
 # Optional: for better performance
 # bitsandbytes>=0.41.0  # for 4-bit/8-bit quantization
 # sentencepiece>=0.1.99  # for some tokenizers