webscout 8.3.6__py3-none-any.whl → 2025.10.11__py3-none-any.whl

webscout/Provider/cerebras.py

@@ -1,340 +1,446 @@
-
-import re
-
-# Import trio before curl_cffi to prevent eventlet socket monkey-patching conflicts
-# See: https://github.com/python-trio/trio/issues/3015
-try:
-    import trio  # noqa: F401
-except ImportError:
-    pass  # trio is optional, ignore if not available
-import json
-from typing import Any, Dict, Generator, List, Optional, Union
-
-import curl_cffi
-from curl_cffi.requests import Session
-
-from webscout import exceptions
-from webscout.AIbase import Provider
-from webscout.AIutel import (  # Import sanitize_stream
-    AwesomePrompts,
-    Conversation,
-    Optimizers,
-    sanitize_stream,
-)
-from webscout.litagent import LitAgent as UserAgent
-
-
-class Cerebras(Provider):
-    """
-    A class to interact with the Cerebras API using a cookie for authentication.
-    """
-
-    AVAILABLE_MODELS = [
-        "qwen-3-coder-480b",
-        "qwen-3-235b-a22b-instruct-2507",
-        "qwen-3-235b-a22b-thinking-2507",
-        "qwen-3-32b",
-        "llama-3.3-70b",
-        "llama-4-maverick-17b-128e-instruct"
-    ]
-
-    def __init__(
-        self,
-        cookie_path: str = None,
-        is_conversation: bool = True,
-        max_tokens: int = 40000,
-        timeout: int = 30,
-        intro: str = None,
-        filepath: str = None,
-        update_file: bool = True,
-        proxies: dict = {},
-        history_offset: int = 10250,
-        act: str = None,
-        api_key: str = None,
-        model: str = "qwen-3-coder-480b",
-        system_prompt: str = "You are a helpful assistant.",
-        temperature: float = 0.7,
-        top_p: float = 0.8,
-    ):
-        # Validate model choice
-        if model not in self.AVAILABLE_MODELS:
-            raise ValueError(
-                f"Invalid model: {model}. Choose from: {self.AVAILABLE_MODELS}"
-            )
-
-        # Initialize basic settings first
-        self.timeout = timeout
-        self.model = model
-        self.system_prompt = system_prompt
-        self.is_conversation = is_conversation
-        self.max_tokens_to_sample = max_tokens
-        self.temperature = temperature
-        self.top_p = top_p
-        self.last_response = {}
-
-        self.session = Session() # Initialize curl_cffi session
-
-        # Handle API key - either provided directly or retrieved from cookies
-        if api_key:
-            self.api_key = api_key.strip()
-            # Basic validation for API key format
-            if not self.api_key or len(self.api_key) < 10:
-                raise ValueError("Invalid API key format. API key must be at least 10 characters long.")
-        elif cookie_path:
-            # Get API key from cookies
-            try:
-                self.api_key = self.get_demo_api_key(cookie_path)
-            except Exception as e:
-                raise exceptions.APIConnectionError(f"Failed to initialize Cerebras client: {e}")
-        else:
-            raise ValueError("Either api_key must be provided or cookie_path must be specified")
-
-        # Initialize optimizers
-        self.__available_optimizers = (
-            method
-            for method in dir(Optimizers)
-            if callable(getattr(Optimizers, method)) and not method.startswith("__")
-        )
-
-        # Initialize conversation settings
-        Conversation.intro = (
-            AwesomePrompts().get_act(
-                act, raise_not_found=True, default="You are a helpful assistant.", case_insensitive=True
-            )
-            if act
-            else "You are a helpful assistant."
-        )
-        self.conversation = Conversation(
-            is_conversation, self.max_tokens_to_sample, filepath, update_file
-        )
-        self.conversation.history_offset = history_offset
-
-        # Apply proxies to the session
-        self.session.proxies = proxies
-
-    # Rest of the class implementation remains the same...
-    @staticmethod
-    def extract_query(text: str) -> str:
-        """Extracts the first code block from the given text."""
-        pattern = r"```(.*?)```"
-        matches = re.findall(pattern, text, re.DOTALL)
-        return matches[0].strip() if matches else text.strip()
-
-    @staticmethod
-    def refiner(text: str) -> str:
-        """Refines the input text by removing surrounding quotes."""
-        return text.strip('"')
-
-    @staticmethod
-    def _cerebras_extractor(chunk: Union[str, Dict[str, Any]]) -> Optional[str]:
-        """Extracts content from Cerebras stream JSON objects."""
-        if isinstance(chunk, dict):
-            return chunk.get("choices", [{}])[0].get("delta", {}).get("content")
-        return None
-
-    def get_demo_api_key(self, cookie_path: str = None) -> str: # Keep this using requests or switch to curl_cffi
-        """Retrieves the demo API key using the provided cookie."""
-        if not cookie_path:
-            raise ValueError("cookie_path must be provided when using cookie-based authentication")
-        try:
-            with open(cookie_path, "r") as file:
-                cookies = {item["name"]: item["value"] for item in json.load(file)}
-        except FileNotFoundError:
-            raise FileNotFoundError(f"Cookie file not found at path: {cookie_path}")
-        except json.JSONDecodeError:
-            raise json.JSONDecodeError("Invalid JSON format in the cookie file.", "", 0)
-
-        headers = {
-            "Accept": "*/*",
-            "Accept-Language": "en-US,en;q=0.9",
-            "Content-Type": "application/json",
-            "Origin": "https://inference.cerebras.ai",
-            "Referer": "https://inference.cerebras.ai/",
-            "user-agent": UserAgent().random(),
-        }
-
-        json_data = {
-            "operationName": "GetMyDemoApiKey",
-            "variables": {},
-            "query": "query GetMyDemoApiKey {\n  GetMyDemoApiKey\n}",
-        }
-
-        try:
-            # Use the initialized curl_cffi session
-            response = self.session.post(
-                "https://inference.cerebras.ai/api/graphql",
-                cookies=cookies,
-                headers=headers,
-                json=json_data,
-                timeout=self.timeout,
-                impersonate="chrome120" # Add impersonate
-            )
-            response.raise_for_status()
-            api_key = response.json().get("data", {}).get("GetMyDemoApiKey")
-            return api_key
-        except curl_cffi.CurlError as e:
-            raise exceptions.APIConnectionError(f"Failed to retrieve API key: {e}")
-        except KeyError:
-            raise exceptions.InvalidResponseError("API key not found in response.")
-
-    def _make_request(self, messages: List[Dict], stream: bool = False) -> Union[Dict, Generator]:
-        """Make a request to the Cerebras API."""
-        headers = {
-            "Authorization": f"Bearer {self.api_key}",
-            "Content-Type": "application/json",
-            "User-Agent": UserAgent().random(),
-        }
-
-        payload = {
-            "model": self.model,
-            "messages": messages,
-            "stream": stream,
-            "max_tokens": self.max_tokens_to_sample,
-            "temperature": self.temperature,
-            "top_p": self.top_p
-        }
-
-        try:
-            # Use the initialized curl_cffi session
-            response = self.session.post(
-                "https://api.cerebras.ai/v1/chat/completions",
-                headers=headers,
-                json=payload,
-                stream=stream,
-                timeout=self.timeout,
-                impersonate="chrome120" # Add impersonate
-            )
-            response.raise_for_status()
-
-            if stream:
-                def generate_stream():
-                    # Use sanitize_stream
-                    processed_stream = sanitize_stream(
-                        data=response.iter_content(chunk_size=None), # Pass byte iterator
-                        intro_value="data:",
-                        to_json=True,     # Stream sends JSON
-                        content_extractor=self._cerebras_extractor, # Use the specific extractor
-                        yield_raw_on_error=False # Skip non-JSON lines or lines where extractor fails
-                    )
-                    for content_chunk in processed_stream:
-                        if content_chunk and isinstance(content_chunk, str):
-                            yield content_chunk # Yield the extracted text chunk
-
-                return generate_stream()
-            else:
-                response_json = response.json()
-                # Extract content for non-streaming response
-                content = response_json.get("choices", [{}])[0].get("message", {}).get("content")
-                return content if content else "" # Return empty string if not found
-
-        except curl_cffi.CurlError as e:
-            raise exceptions.APIConnectionError(f"Request failed (CurlError): {e}") from e
-        except Exception as e:
-            # Check if it's an HTTP error with status code
-            if hasattr(e, 'response') and hasattr(e.response, 'status_code'):
-                status_code = e.response.status_code
-                if status_code == 401:
-                    raise exceptions.APIConnectionError(
-                        "Authentication failed (401): Invalid API key. Please check your API key and try again."
-                    ) from e
-                elif status_code == 403:
-                    raise exceptions.APIConnectionError(
-                        "Access forbidden (403): Your API key may not have permission to access this resource."
-                    ) from e
-                elif status_code == 429:
-                    raise exceptions.APIConnectionError(
-                        "Rate limit exceeded (429): Too many requests. Please wait and try again."
-                    ) from e
-                else:
-                    raise exceptions.APIConnectionError(f"HTTP {status_code} error: {e}") from e
-            else:
-                raise exceptions.APIConnectionError(f"Request failed: {e}") from e
-
-    def ask(
-        self,
-        prompt: str,
-        stream: bool = False,
-        raw: bool = False, # Add raw parameter for consistency
-        optimizer: str = None,
-        conversationally: bool = False,
-    ) -> Union[Dict, Generator]:
-        """Send a prompt to the model and get a response."""
-        conversation_prompt = self.conversation.gen_complete_prompt(prompt)
-        if optimizer:
-            if optimizer in self.__available_optimizers:
-                conversation_prompt = getattr(Optimizers, optimizer)(
-                    conversation_prompt if conversationally else prompt
-                )
-            else:
-                raise Exception(f"Optimizer is not one of {self.__available_optimizers}")
-
-        messages = [
-            {"role": "system", "content": self.system_prompt},
-            {"role": "user", "content": conversation_prompt}
-        ]
-
-        try:
-            response = self._make_request(messages, stream)
-
-            if stream:
-                # Wrap the generator to yield dicts or raw strings
-                def stream_wrapper():
-                    full_text = ""
-                    for chunk in response:
-                        full_text += chunk
-                        yield chunk if raw else {"text": chunk}
-                    # Update history after stream finishes
-                    self.last_response = {"text": full_text}
-                    self.conversation.update_chat_history(prompt, full_text)
-                return stream_wrapper()
-            else:
-                # Non-streaming response is already the full text string
-                self.last_response = {"text": response}
-                self.conversation.update_chat_history(prompt, response)
-                return self.last_response if not raw else response # Return dict or raw string
-
-        except Exception as e:
-            raise exceptions.FailedToGenerateResponseError(f"Error during request: {e}")
-
-    def chat(
-        self,
-        prompt: str,
-        stream: bool = False,
-        optimizer: str = None,
-        conversationally: bool = False,
-    ) -> Union[str, Generator]:
-        """Chat with the model."""
-        # Ask returns a generator for stream=True, dict/str for stream=False
-        response_gen_or_dict = self.ask(prompt, stream, raw=False, optimizer=optimizer, conversationally=conversationally)
-
-        if stream:
-            # Wrap the generator from ask() to get message text
-            def stream_wrapper():
-                for chunk_dict in response_gen_or_dict:
-                    yield self.get_message(chunk_dict)
-            return stream_wrapper()
-        else:
-            # Non-streaming response is already a dict
-            return self.get_message(response_gen_or_dict)
-
-    def get_message(self, response: str) -> str:
-        """Retrieves message from response."""
-        # Updated to handle dict input from ask()
-        assert isinstance(response, dict), "Response should be of dict data-type only for get_message"
-        return response.get("text", "")
-
-
-if __name__ == "__main__":
-    from rich import print
-
-    # Example usage
-    cerebras = Cerebras(
-        api_key='csk-**********************',  # Replace with your actual API key
-        model='qwen-3-235b-a22b-instruct-2507',
-        system_prompt="You are a helpful AI assistant."
-    )
-
-    # Test with streaming
-    response = cerebras.chat("Hello!", stream=True)
-    for chunk in response:
-        print(chunk, end="", flush=True)
+import re
+
+
+# Import trio before curl_cffi to prevent eventlet socket monkey-patching conflicts
+# See: https://github.com/python-trio/trio/issues/3015
+try:
+    import trio  # noqa: F401
+except ImportError:
+    pass  # trio is optional, ignore if not available
+import json
+from typing import Any, Dict, Generator, List, Optional, Union
+
+import curl_cffi
+from curl_cffi.requests import Session
+
+from webscout import exceptions
+from webscout.AIbase import Provider
+from webscout.AIutel import (  # Import sanitize_stream
+    AwesomePrompts,
+    Conversation,
+    Optimizers,
+    sanitize_stream,
+)
+from webscout.litagent import LitAgent as UserAgent
+
+
+class Cerebras(Provider):
+    """
+    A class to interact with the Cerebras API using a cookie for authentication.
+    """
+    required_auth = True
+    AVAILABLE_MODELS = [
+        "qwen-3-coder-480b",
+        "qwen-3-235b-a22b-instruct-2507",
+        "qwen-3-235b-a22b-thinking-2507",
+        "qwen-3-32b",
+        "llama-3.3-70b",
+        "llama-4-maverick-17b-128e-instruct",
+        "gpt-oss-120b",
+        "llama-4-scout-17b-16e-instruct",
+        "llama3.1-8b"
+    ]
+
+    @classmethod
+    def get_models(cls, api_key: str = None):
+        """Fetch available models from Cerebras API.
+        
+        Args:
+            api_key (str, optional): Cerebras API key. If not provided, returns default models.
+            
+        Returns:
+            list: List of available model IDs
+        """
+        if not api_key:
+            return cls.AVAILABLE_MODELS
+            
+        try:
+            # Use a temporary curl_cffi session for this class method
+            temp_session = Session()
+            headers = {
+                "Content-Type": "application/json",
+                "Authorization": f"Bearer {api_key}",
+            }
+            
+            response = temp_session.get(
+                "https://api.cerebras.ai/v1/models",
+                headers=headers,
+                impersonate="chrome120"
+            )
+            
+            if response.status_code != 200:
+                return cls.AVAILABLE_MODELS
+                
+            data = response.json()
+            if "data" in data and isinstance(data["data"], list):
+                return [model['id'] for model in data['data']]
+            return cls.AVAILABLE_MODELS
+            
+        except Exception:
+            # Fallback to default models list if fetching fails
+            return cls.AVAILABLE_MODELS
+
+    @classmethod
+    def update_available_models(cls, api_key=None):
+        """Update the available models list from Cerebras API"""
+        try:
+            models = cls.get_models(api_key)
+            if models and len(models) > 0:
+                cls.AVAILABLE_MODELS = models
+        except Exception:
+            # Fallback to default models list if fetching fails
+            pass
+
+    def __init__(
+        self,
+        cookie_path: str = None,
+        is_conversation: bool = True,
+        max_tokens: int = 40000,
+        timeout: int = 30,
+        intro: str = None,
+        filepath: str = None,
+        update_file: bool = True,
+        proxies: dict = {},
+        history_offset: int = 10250,
+        act: str = None,
+        api_key: str = None,
+        model: str = "qwen-3-coder-480b",
+        system_prompt: str = "You are a helpful assistant.",
+        temperature: float = 0.7,
+        top_p: float = 0.8,
+    ):
+        # Initialize basic settings first
+        self.timeout = timeout
+        self.model = model
+        self.system_prompt = system_prompt
+        self.is_conversation = is_conversation
+        self.max_tokens_to_sample = max_tokens
+        self.temperature = temperature
+        self.top_p = top_p
+        self.last_response = {}
+
+        self.session = Session() # Initialize curl_cffi session
+
+        # Handle API key - either provided directly or retrieved from cookies
+        if api_key:
+            self.api_key = api_key.strip()
+            # Basic validation for API key format
+            if not self.api_key or len(self.api_key) < 10:
+                raise ValueError("Invalid API key format. API key must be at least 10 characters long.")
+        elif cookie_path:
+            # Get API key from cookies
+            try:
+                self.api_key = self.get_demo_api_key(cookie_path)
+            except Exception as e:
+                raise exceptions.APIConnectionError(f"Failed to initialize Cerebras client: {e}")
+        else:
+            raise ValueError("Either api_key must be provided or cookie_path must be specified")
+
+        # Update available models from API
+        self.update_available_models(self.api_key)
+
+        # Validate model choice after updating models
+        if model not in self.AVAILABLE_MODELS:
+            raise ValueError(
+                f"Invalid model: {model}. Choose from: {self.AVAILABLE_MODELS}"
+            )
+
+        # Initialize optimizers
+        self.__available_optimizers = (
+            method
+            for method in dir(Optimizers)
+            if callable(getattr(Optimizers, method)) and not method.startswith("__")
+        )
+
+        # Initialize conversation settings
+        Conversation.intro = (
+            AwesomePrompts().get_act(
+                act, raise_not_found=True, default="You are a helpful assistant.", case_insensitive=True
+            )
+            if act
+            else "You are a helpful assistant."
+        )
+        self.conversation = Conversation(
+            is_conversation, self.max_tokens_to_sample, filepath, update_file
+        )
+        self.conversation.history_offset = history_offset
+
+        # Apply proxies to the session
+        self.session.proxies = proxies
+
+    @classmethod
+    def get_models(cls, api_key: str = None):
+        """Fetch available models from Cerebras API.
+        
+        Args:
+            api_key (str, optional): Cerebras API key. If not provided, returns default models.
+            
+        Returns:
+            list: List of available model IDs
+        """
+        if not api_key:
+            return cls.AVAILABLE_MODELS
+            
+        try:
+            # Use a temporary curl_cffi session for this class method
+            temp_session = Session()
+            headers = {
+                "Content-Type": "application/json",
+                "Authorization": f"Bearer {api_key}",
+            }
+            
+            response = temp_session.get(
+                "https://api.cerebras.ai/v1/models",
+                headers=headers,
+                impersonate="chrome120"
+            )
+            
+            if response.status_code != 200:
+                return cls.AVAILABLE_MODELS
+                
+            data = response.json()
+            if "data" in data and isinstance(data["data"], list):
+                return [model['id'] for model in data['data']]
+            return cls.AVAILABLE_MODELS
+            
+        except Exception:
+            # Fallback to default models list if fetching fails
+            return cls.AVAILABLE_MODELS
+
+    @classmethod
+    def update_available_models(cls, api_key=None):
+        """Update the available models list from Cerebras API"""
+        try:
+            models = cls.get_models(api_key)
+            if models and len(models) > 0:
+                cls.AVAILABLE_MODELS = models
+        except Exception:
+            # Fallback to default models list if fetching fails
+            pass
+
+    # Rest of the class implementation remains the same...
+    @staticmethod
+    def extract_query(text: str) -> str:
+        """Extracts the first code block from the given text."""
+        pattern = r"```(.*?)```"
+        matches = re.findall(pattern, text, re.DOTALL)
+        return matches[0].strip() if matches else text.strip()
+
+    @staticmethod
+    def refiner(text: str) -> str:
+        """Refines the input text by removing surrounding quotes."""
+        return text.strip('"')
+
+    @staticmethod
+    def _cerebras_extractor(chunk: Union[str, Dict[str, Any]]) -> Optional[str]:
+        """Extracts content from Cerebras stream JSON objects."""
+        if isinstance(chunk, dict):
+            return chunk.get("choices", [{}])[0].get("delta", {}).get("content")
+        return None
+
+    def get_demo_api_key(self, cookie_path: str = None) -> str: # Keep this using requests or switch to curl_cffi
+        """Retrieves the demo API key using the provided cookie."""
+        if not cookie_path:
+            raise ValueError("cookie_path must be provided when using cookie-based authentication")
+        try:
+            with open(cookie_path, "r") as file:
+                cookies = {item["name"]: item["value"] for item in json.load(file)}
+        except FileNotFoundError:
+            raise FileNotFoundError(f"Cookie file not found at path: {cookie_path}")
+        except json.JSONDecodeError:
+            raise json.JSONDecodeError("Invalid JSON format in the cookie file.", "", 0)
+
+        headers = {
+            "Accept": "*/*",
+            "Accept-Language": "en-US,en;q=0.9",
+            "Content-Type": "application/json",
+            "Origin": "https://inference.cerebras.ai",
+            "Referer": "https://inference.cerebras.ai/",
+            "user-agent": UserAgent().random(),
+        }
+
+        json_data = {
+            "operationName": "GetMyDemoApiKey",
+            "variables": {},
+            "query": "query GetMyDemoApiKey {\n  GetMyDemoApiKey\n}",
+        }
+
+        try:
+            # Use the initialized curl_cffi session
+            response = self.session.post(
+                "https://inference.cerebras.ai/api/graphql",
+                cookies=cookies,
+                headers=headers,
+                json=json_data,
+                timeout=self.timeout,
+                impersonate="chrome120" # Add impersonate
+            )
+            response.raise_for_status()
+            api_key = response.json().get("data", {}).get("GetMyDemoApiKey")
+            return api_key
+        except curl_cffi.CurlError as e:
+            raise exceptions.APIConnectionError(f"Failed to retrieve API key: {e}")
+        except KeyError:
+            raise exceptions.InvalidResponseError("API key not found in response.")
+
+    def _make_request(self, messages: List[Dict], stream: bool = False) -> Union[Dict, Generator]:
+        """Make a request to the Cerebras API."""
+        headers = {
+            "Authorization": f"Bearer {self.api_key}",
+            "Content-Type": "application/json",
+            "User-Agent": UserAgent().random(),
+        }
+
+        payload = {
+            "model": self.model,
+            "messages": messages,
+            "stream": stream,
+            "max_tokens": self.max_tokens_to_sample,
+            "temperature": self.temperature,
+            "top_p": self.top_p
+        }
+
+        try:
+            # Use the initialized curl_cffi session
+            response = self.session.post(
+                "https://api.cerebras.ai/v1/chat/completions",
+                headers=headers,
+                json=payload,
+                stream=stream,
+                timeout=self.timeout,
+                impersonate="chrome120" # Add impersonate
+            )
+            response.raise_for_status()
+
+            if stream:
+                def generate_stream():
+                    # Use sanitize_stream
+                    processed_stream = sanitize_stream(
+                        data=response.iter_content(chunk_size=None), # Pass byte iterator
+                        intro_value="data:",
+                        to_json=True,     # Stream sends JSON
+                        content_extractor=self._cerebras_extractor, # Use the specific extractor
+                        yield_raw_on_error=False # Skip non-JSON lines or lines where extractor fails
+                    )
+                    for content_chunk in processed_stream:
+                        if content_chunk and isinstance(content_chunk, str):
+                            yield content_chunk # Yield the extracted text chunk
+
+                return generate_stream()
+            else:
+                response_json = response.json()
+                # Extract content for non-streaming response
+                content = response_json.get("choices", [{}])[0].get("message", {}).get("content")
+                return content if content else "" # Return empty string if not found
+
+        except curl_cffi.CurlError as e:
+            raise exceptions.APIConnectionError(f"Request failed (CurlError): {e}") from e
+        except Exception as e:
+            # Check if it's an HTTP error with status code
+            if hasattr(e, 'response') and hasattr(e.response, 'status_code'):
+                status_code = e.response.status_code
+                if status_code == 401:
+                    raise exceptions.APIConnectionError(
+                        "Authentication failed (401): Invalid API key. Please check your API key and try again."
+                    ) from e
+                elif status_code == 403:
+                    raise exceptions.APIConnectionError(
+                        "Access forbidden (403): Your API key may not have permission to access this resource."
+                    ) from e
+                elif status_code == 429:
+                    raise exceptions.APIConnectionError(
+                        "Rate limit exceeded (429): Too many requests. Please wait and try again."
+                    ) from e
+                else:
+                    raise exceptions.APIConnectionError(f"HTTP {status_code} error: {e}") from e
+            else:
+                raise exceptions.APIConnectionError(f"Request failed: {e}") from e
+
+    def ask(
+        self,
+        prompt: str,
+        stream: bool = False,
+        raw: bool = False, # Add raw parameter for consistency
+        optimizer: str = None,
+        conversationally: bool = False,
+    ) -> Union[Dict, Generator]:
+        """Send a prompt to the model and get a response."""
+        conversation_prompt = self.conversation.gen_complete_prompt(prompt)
+        if optimizer:
+            if optimizer in self.__available_optimizers:
+                conversation_prompt = getattr(Optimizers, optimizer)(
+                    conversation_prompt if conversationally else prompt
+                )
+            else:
+                raise Exception(f"Optimizer is not one of {self.__available_optimizers}")
+
+        messages = [
+            {"role": "system", "content": self.system_prompt},
+            {"role": "user", "content": conversation_prompt}
+        ]
+
+        try:
+            response = self._make_request(messages, stream)
+
+            if stream:
+                # Wrap the generator to yield dicts or raw strings
+                def stream_wrapper():
+                    full_text = ""
+                    for chunk in response:
+                        full_text += chunk
+                        yield chunk if raw else {"text": chunk}
+                    # Update history after stream finishes
+                    self.last_response = {"text": full_text}
+                    self.conversation.update_chat_history(prompt, full_text)
+                return stream_wrapper()
+            else:
+                # Non-streaming response is already the full text string
+                self.last_response = {"text": response}
+                self.conversation.update_chat_history(prompt, response)
+                return self.last_response if not raw else response # Return dict or raw string
+
+        except Exception as e:
+            raise exceptions.FailedToGenerateResponseError(f"Error during request: {e}")
+
+    def chat(
+        self,
+        prompt: str,
+        stream: bool = False,
+        optimizer: str = None,
+        conversationally: bool = False,
+    ) -> Union[str, Generator]:
+        """Chat with the model."""
+        # Ask returns a generator for stream=True, dict/str for stream=False
+        response_gen_or_dict = self.ask(prompt, stream, raw=False, optimizer=optimizer, conversationally=conversationally)
+
+        if stream:
+            # Wrap the generator from ask() to get message text
+            def stream_wrapper():
+                for chunk_dict in response_gen_or_dict:
+                    yield self.get_message(chunk_dict)
+            return stream_wrapper()
+        else:
+            # Non-streaming response is already a dict
+            return self.get_message(response_gen_or_dict)
+
+    def get_message(self, response: str) -> str:
+        """Retrieves message from response."""
+        # Updated to handle dict input from ask()
+        assert isinstance(response, dict), "Response should be of dict data-type only for get_message"
+        return response.get("text", "")
+
+
+if __name__ == "__main__":
+    from rich import print
+
+    # Example usage
+    cerebras = Cerebras(
+        api_key='csk-**********************',  # Replace with your actual API key
+        model='qwen-3-235b-a22b-instruct-2507',
+        system_prompt="You are a helpful AI assistant."
+    )
+
+    # Test with streaming
+    response = cerebras.chat("Hello!", stream=True)
+    for chunk in response:
+        print(chunk, end="", flush=True)