webscout 8.3.2__py3-none-any.whl → 8.3.3__py3-none-any.whl

webscout/Provider/Nemotron.py

@@ -110,7 +110,8 @@ class NEMOTRON(Provider):
     def _make_request(
         self,
         message: str,
-        stream: bool = False
+        stream: bool = False,
+        raw: bool = False
     ) -> Generator[str, None, None]:
         """Make request to NEMOTRON API."""
         payload = {
@@ -131,10 +132,26 @@ class NEMOTRON(Provider):
                     timeout=self.timeout
                 ) as response:
                     response.raise_for_status()
-                    yield from sanitize_stream(
-                        response.iter_content(chunk_size=1024),
-                        to_json=False,
-                    )
+                    buffer = ""
+                    chunk_size = 32
+                    for chunk in response.iter_content(chunk_size=chunk_size):
+                        if not chunk:
+                            continue
+                        text = chunk.decode(errors="ignore")
+                        buffer += text
+                        while len(buffer) >= chunk_size:
+                            out = buffer[:chunk_size]
+                            buffer = buffer[chunk_size:]
+                            if out.strip():
+                                if raw:
+                                    yield out
+                                else:
+                                    yield out
+                    if buffer.strip():
+                        if raw:
+                            yield buffer
+                        else:
+                            yield buffer
             else:
                 response = self.session.post(
                     self.url,
@@ -143,7 +160,10 @@ class NEMOTRON(Provider):
                     timeout=self.timeout
                 )
                 response.raise_for_status()
-                yield response.text
+                if raw:
+                    yield response.text
+                else:
+                    yield response.text
 
         except requests.exceptions.RequestException as e:
             raise exceptions.ProviderConnectionError(f"Connection error: {str(e)}")
@@ -167,13 +187,20 @@ class NEMOTRON(Provider):
                 raise ValueError(f"Optimizer is not one of {self.__available_optimizers}")
 
         def for_stream():
-            for text in self._make_request(conversation_prompt, stream=True):
-                yield {"text": text}
+            for text in self._make_request(conversation_prompt, stream=True, raw=raw):
+                if raw:
+                    yield text
+                else:
+                    yield {"text": text}
 
         def for_non_stream():
-            response_text = next(self._make_request(conversation_prompt, stream=False))
-            self.last_response = {"text": response_text}
-            return self.last_response
+            response_text = next(self._make_request(conversation_prompt, stream=False, raw=raw))
+            if raw:
+                self.last_response = response_text
+                return response_text
+            else:
+                self.last_response = {"text": response_text}
+                return self.last_response
 
         return for_stream() if stream else for_non_stream()
 
@@ -214,5 +241,6 @@ class NEMOTRON(Provider):
 if __name__ == "__main__":
     # Example usage
     nemotron = NEMOTRON()
-    response = nemotron.chat("Hello, how are you?", stream=False)
-    print(response)
+    response = nemotron.chat("write me about humans in points", stream=True)
+    for part in response:
+        print(part, end="", flush=True)

webscout/Provider/Netwrck.py

@@ -127,75 +127,58 @@ class Netwrck(Provider):
 
         def for_stream():
             try:
-                # Use curl_cffi session post with impersonate
                 response = self.session.post(
                     "https://netwrck.com/api/chatpred_or",
                     json=payload,
-                    # headers are set on the session
-                    # proxies are set on the session
                     timeout=self.timeout,
                     stream=True,
-                    impersonate="chrome110" # Use a common impersonation profile
+                    impersonate="chrome110"
                 )
-                response.raise_for_status() # Check for HTTP errors
-
-                streaming_text = ""
-                # Use sanitize_stream
-                processed_stream = sanitize_stream(
-                    data=response.iter_content(chunk_size=None), # Pass byte iterator
-                    intro_value=None, # No prefix
-                    to_json=False,    # It's text
-                    content_extractor=self._netwrck_extractor, # Use the quote stripper
-                    yield_raw_on_error=True
-                )
-                for content_chunk in processed_stream:
-                    if content_chunk and isinstance(content_chunk, str):
-                        streaming_text += content_chunk
-                        yield {"text": content_chunk} if not raw else content_chunk
-                # Update history after stream finishes
-                self.last_response = {"text": streaming_text} # Store aggregated text
-                self.conversation.update_chat_history(payload["query"], streaming_text)
-
-            except CurlError as e: # Catch CurlError
+                response.raise_for_status()
+                buffer = ""
+                chunk_size = 32
+                for chunk in response.iter_content(chunk_size=chunk_size):
+                    if not chunk:
+                        continue
+                    text = chunk.decode(errors="ignore")
+                    buffer += text
+                    while len(buffer) >= chunk_size:
+                        out = buffer[:chunk_size]
+                        buffer = buffer[chunk_size:]
+                        if out.strip():
+                            if raw:
+                                yield out
+                            else:
+                                yield {"text": out}
+                if buffer.strip():
+                    if raw:
+                        yield buffer
+                    else:
+                        yield {"text": buffer}
+                self.last_response = {"text": buffer}
+                self.conversation.update_chat_history(payload["query"], buffer)
+            except CurlError as e:
                 raise exceptions.ProviderConnectionError(f"Network error (CurlError): {str(e)}") from e
-            except Exception as e: # Catch other potential exceptions (like HTTPError)
+            except Exception as e:
                 err_text = getattr(e, 'response', None) and getattr(e.response, 'text', '')
                 raise exceptions.ProviderConnectionError(f"Unexpected error ({type(e).__name__}): {str(e)} - {err_text}") from e
 
         def for_non_stream():
             try:
-                # Use curl_cffi session post with impersonate
                 response = self.session.post(
                     "https://netwrck.com/api/chatpred_or",
                     json=payload,
-                    # headers are set on the session
-                    # proxies are set on the session
                     timeout=self.timeout,
-                    impersonate="chrome110" # Use a common impersonation profile
-                )
-                response.raise_for_status() # Check for HTTP errors
-                
-                response_text_raw = response.text # Get raw text
-
-                # Process the text using sanitize_stream
-                processed_stream = sanitize_stream(
-                    data=response_text_raw,
-                    intro_value=None,
-                    to_json=False,
-                    content_extractor=self._netwrck_extractor
+                    impersonate="chrome110"
                 )
-                # Aggregate the single result
-                text = "".join(list(processed_stream))
-
-                self.last_response = {"text": text} # Store processed text
-                self.conversation.update_chat_history(prompt, text)
-
-                # Return dict or raw string
-                return text if raw else self.last_response
-
-            except CurlError as e: # Catch CurlError
+                response.raise_for_status()
+                response_text_raw = response.text
+                self.last_response = {"text": response_text_raw}
+                self.conversation.update_chat_history(prompt, response_text_raw)
+                return response_text_raw if raw else self.last_response
+            except CurlError as e:
                 raise exceptions.FailedToGenerateResponseError(f"Network error (CurlError): {str(e)}") from e
-            except Exception as e: # Catch other potential exceptions (like HTTPError)
+            except Exception as e:
                 err_text = getattr(e, 'response', None) and getattr(e.response, 'text', '')
                 raise exceptions.FailedToGenerateResponseError(f"Unexpected error ({type(e).__name__}): {str(e)} - {err_text}") from e
 

webscout/Provider/OPENAI/BLACKBOXAI.py

@@ -1043,4 +1043,3 @@ if __name__ == "__main__":
     )
     for chunk in response:
         print(chunk.choices[0].delta.content, end='', flush=True)
-

webscout/Provider/OPENAI/MiniMax.py

@@ -0,0 +1,298 @@
+import os
+import requests
+import json
+import time
+import uuid
+from typing import List, Dict, Optional, Union, Generator, Any
+
+from webscout.Provider.OPENAI.base import OpenAICompatibleProvider, BaseChat, BaseCompletions
+from webscout.Provider.OPENAI.utils import (
+    ChatCompletionChunk, ChatCompletion, Choice, ChoiceDelta,
+    ChatCompletionMessage, CompletionUsage, count_tokens
+)
+
+class Completions(BaseCompletions):
+    def __init__(self, client: 'MiniMax'):
+        self._client = client
+
+    def create(
+        self,
+        *,
+        model: str,
+        messages: List[Dict[str, str]],
+        max_tokens: Optional[int] = None,
+        stream: bool = False,
+        temperature: Optional[float] = None,
+        top_p: Optional[float] = None,
+        timeout: Optional[int] = None,
+        proxies: Optional[Dict[str, str]] = None,
+        stop: Optional[Union[str, List[str]]] = None,
+        **kwargs: Any
+    ) -> Union[ChatCompletion, Generator[ChatCompletionChunk, None, None]]:
+        """
+        Creates a model response for the given chat conversation.
+        Mimics openai.chat.completions.create
+        """
+        api_key = self._client.api_key
+        if not api_key:
+            raise Exception("MINIMAX_API_KEY not set in environment.")
+        model_name = self._client.convert_model_name(model)
+        payload = {
+            "model": model_name,
+            "messages": messages,
+            "stream": stream,
+        }
+        if max_tokens is not None:
+            payload["max_tokens"] = max_tokens
+        if temperature is not None:
+            payload["temperature"] = temperature
+        if top_p is not None:
+            payload["top_p"] = top_p
+        if stop is not None:
+            payload["stop"] = stop
+        payload.update(kwargs)
+        request_id = f"chatcmpl-{uuid.uuid4()}"
+        created_time = int(time.time())
+        if stream:
+            return self._create_stream(request_id, created_time, model_name, payload, timeout, proxies)
+        else:
+            return self._create_non_stream(request_id, created_time, model_name, payload, timeout, proxies)
+
+    def _create_stream(
+        self, request_id: str, created_time: int, model: str, payload: Dict[str, Any], timeout: Optional[int] = None, proxies: Optional[Dict[str, str]] = None
+    ) -> Generator[ChatCompletionChunk, None, None]:
+        try:
+            headers = {
+                'Content-Type': 'application/json',
+                'Authorization': f'Bearer {self._client.api_key}',
+            }
+            response = self._client.session.post(
+                self._client.api_endpoint,
+                headers=headers,
+                data=json.dumps(payload),
+                stream=True,
+                timeout=timeout or self._client.timeout,
+                proxies=proxies
+            )
+            response.raise_for_status()
+            prompt_tokens = count_tokens([msg.get("content", "") for msg in payload.get("messages", [])])
+            completion_tokens = 0
+            total_tokens = prompt_tokens
+            streaming_response = ""
+            last_content = ""
+            last_reasoning = ""
+            in_think = False
+            for line in response.iter_lines():
+                if line:
+                    line = line.decode('utf-8')
+                    if line.startswith('data: '):
+                        line = line[6:]
+                        if line.strip() == '[DONE]':
+                            break
+                        try:
+                            chunk_data = json.loads(line)
+                            if 'choices' in chunk_data and chunk_data['choices']:
+                                choice_data = chunk_data['choices'][0]
+                                delta = choice_data.get('delta', {})
+                                content = delta.get('content')
+                                reasoning_content = delta.get('reasoning_content')
+                                finish_reason = choice_data.get('finish_reason')
+                                # Only yield <think> and reasoning_content if reasoning_content is not empty
+                                if reasoning_content and reasoning_content.strip() and reasoning_content != last_reasoning:
+                                    if not in_think:
+                                        yield ChatCompletionChunk(
+                                            id=request_id,
+                                            choices=[Choice(index=0, delta=ChoiceDelta(content='<think>\n\n', role=None, tool_calls=None), finish_reason=None, logprobs=None)],
+                                            created=created_time,
+                                            model=model
+                                        )
+                                        in_think = True
+                                    yield ChatCompletionChunk(
+                                        id=request_id,
+                                        choices=[Choice(index=0, delta=ChoiceDelta(content=reasoning_content, role=None, tool_calls=None), finish_reason=None, logprobs=None)],
+                                        created=created_time,
+                                        model=model
+                                    )
+                                    last_reasoning = reasoning_content
+                                # Only yield </think> if we were in <think> and now have new content
+                                if in_think and content and content.strip() and content != last_content:
+                                    yield ChatCompletionChunk(
+                                        id=request_id,
+                                        choices=[Choice(index=0, delta=ChoiceDelta(content='</think>\n\n', role=None, tool_calls=None), finish_reason=None, logprobs=None)],
+                                        created=created_time,
+                                        model=model
+                                    )
+                                    in_think = False
+                                # Only yield content if it is not empty
+                                if content and content.strip() and content != last_content:
+                                    completion_tokens += count_tokens(content)
+                                    total_tokens = prompt_tokens + completion_tokens
+                                    choice_delta = ChoiceDelta(
+                                        content=content,
+                                        role=delta.get('role', 'assistant'),
+                                        tool_calls=delta.get('tool_calls')
+                                    )
+                                    choice = Choice(
+                                        index=0,
+                                        delta=choice_delta,
+                                        finish_reason=finish_reason,
+                                        logprobs=None
+                                    )
+                                    chunk = ChatCompletionChunk(
+                                        id=request_id,
+                                        choices=[choice],
+                                        created=created_time,
+                                        model=model
+                                    )
+                                    chunk.usage = {
+                                        "prompt_tokens": prompt_tokens,
+                                        "completion_tokens": completion_tokens,
+                                        "total_tokens": total_tokens,
+                                        "estimated_cost": None
+                                    }
+                                    yield chunk
+                                    streaming_response += content
+                                    last_content = content
+                        except Exception:
+                            continue
+            # Final chunk with finish_reason="stop"
+            delta = ChoiceDelta(content=None, role=None, tool_calls=None)
+            choice = Choice(index=0, delta=delta, finish_reason="stop", logprobs=None)
+            chunk = ChatCompletionChunk(
+                id=request_id,
+                choices=[choice],
+                created=created_time,
+                model=model
+            )
+            chunk.usage = {
+                "prompt_tokens": prompt_tokens,
+                "completion_tokens": completion_tokens,
+                "total_tokens": total_tokens,
+                "estimated_cost": None
+            }
+            yield chunk
+        except Exception as e:
+            raise IOError(f"MiniMax stream request failed: {e}") from e
+
+    def _create_non_stream(
+        self, request_id: str, created_time: int, model: str, payload: Dict[str, Any], timeout: Optional[int] = None, proxies: Optional[Dict[str, str]] = None
+    ) -> ChatCompletion:
+        try:
+            headers = {
+                'Content-Type': 'application/json',
+                'Authorization': f'Bearer {self._client.api_key}',
+            }
+            payload_copy = payload.copy()
+            payload_copy["stream"] = False
+            response = self._client.session.post(
+                self._client.api_endpoint,
+                headers=headers,
+                data=json.dumps(payload_copy),
+                timeout=timeout or self._client.timeout,
+                proxies=proxies
+            )
+            response.raise_for_status()
+            data = response.json()
+            full_text = ""
+            finish_reason = "stop"
+            if 'choices' in data and data['choices']:
+                choice_data = data['choices'][0]
+                # MiniMax returns content in 'message' or directly in 'delta' for streaming
+                reasoning_content = ""
+                if 'message' in choice_data and choice_data['message']:
+                    full_text = choice_data['message'].get('content', '')
+                    reasoning_content = choice_data['message'].get('reasoning_content', '')
+                elif 'delta' in choice_data and choice_data['delta']:
+                    full_text = choice_data['delta'].get('content', '')
+                    reasoning_content = choice_data['delta'].get('reasoning_content', '')
+                finish_reason = choice_data.get('finish_reason', 'stop')
+                # If both are present, concatenate with <think> ... </think>
+                if reasoning_content and reasoning_content.strip():
+                    if full_text and full_text.strip():
+                        full_text = f"<think>\n\n{reasoning_content}</think>\n\n{full_text}"
+                    else:
+                        full_text = f"<think>\n\n{reasoning_content}</think>\n\n"
+            message = ChatCompletionMessage(
+                role="assistant",
+                content=full_text
+            )
+            choice = Choice(
+                index=0,
+                message=message,
+                finish_reason=finish_reason
+            )
+            prompt_tokens = count_tokens([msg.get("content", "") for msg in payload.get("messages", [])])
+            completion_tokens = count_tokens(full_text)
+            usage = CompletionUsage(
+                prompt_tokens=prompt_tokens,
+                completion_tokens=completion_tokens,
+                total_tokens=prompt_tokens + completion_tokens
+            )
+            completion = ChatCompletion(
+                id=request_id,
+                choices=[choice],
+                created=created_time,
+                model=model,
+                usage=usage,
+            )
+            return completion
+        except Exception as e:
+            raise IOError(f"MiniMax non-stream request failed: {e}") from e
+
+class Chat(BaseChat):
+    def __init__(self, client: 'MiniMax'):
+        self.completions = Completions(client)
+
+class MiniMax(OpenAICompatibleProvider):
+    """
+    OpenAI-compatible client for MiniMax API.
+    """
+    AVAILABLE_MODELS = [
+        "MiniMax-Reasoning-01"
+    ]
+    def __init__(self, timeout: int = 30):
+        self.timeout = timeout
+        self.api_endpoint = "https://api.minimaxi.chat/v1/text/chatcompletion_v2"
+        self.session = requests.Session()
+        self.api_key = "eyJhbGciOiJSUzI1NiIsInR5cCI6IkpXVCJ9.eyJHcm91cE5hbWUiOiJtbyBuaSIsIlVzZXJOYW1lIjoibW8gbmkiLCJBY2NvdW50IjoiIiwiU3ViamVjdElEIjoiMTg3NjIwMDY0ODA2NDYzNTI0MiIsIlBob25lIjoiIiwiR3JvdXBJRCI6IjE4NzYyMDA2NDgwNjA0NDA5MzgiLCJQYWdlTmFtZSI6IiIsIk1haWwiOiJuaW1vQHN1YnN1cC52aXAiLCJDcmVhdGVUaW1lIjoiMjAyNS0wMS0wNyAxMToyNzowNyIsIlRva2VuVHlwZSI6MSwiaXNzIjoibWluaW1heCJ9.Ge1ZnpFPUfXVdMini0P_qXbP_9VYwzXiffG9DsNQck4GtYEOs33LDeAiwrVsrrLZfvJ2icQZ4sRZS54wmPuWua_Dav6pYJty8ZtahmUX1IuhlUX5YErhhCRAIy3J1xB8FkLHLyylChuBHpkNz6O6BQLmPqmoa-cOYK9Qrc6IDeu8SX1iMzO9-MSkcWNvkvpCF2Pf9tekBVWNKMDK6IZoMEPbtkaPXdDyP6l0M0e2AlL_E0oM9exg3V-ohAi8OTPFyqM6dcd4TwF-b9DULxfIsRFw401mvIxcTDWa42u2LULewdATVRD2BthU65tuRqEiWeFWMvFlPj2soMze_QIiUA"
+        self.chat = Chat(self)
+
+    @property
+    def models(self):
+        class _ModelList:
+            def list(inner_self):
+                return MiniMax.AVAILABLE_MODELS
+        return _ModelList()
+
+    def convert_model_name(self, model: str) -> str:
+        if model in self.AVAILABLE_MODELS:
+            return model
+        return self.AVAILABLE_MODELS[0]
+
+if __name__ == "__main__":
+    from rich import print
+    client = MiniMax()
+    messages = [
+        {"role": "user", "content": "What is the capital of France?"}
+    ]
+    # Non-streaming example
+    response = client.chat.completions.create(
+        model="MiniMax-Reasoning-01",
+        messages=messages,
+        max_tokens=5000,
+        stream=False
+    )
+    print("Non-streaming response:")
+    print(response)
+    # Streaming example
+    print("\nStreaming response:")
+    stream = client.chat.completions.create(
+        model="MiniMax-Reasoning-01",
+        messages=messages,
+        max_tokens=5000,
+        stream=True
+    )
+    for chunk in stream:
+        if chunk.choices[0].delta and chunk.choices[0].delta.content:
+            print(chunk.choices[0].delta.content, end="")
+    print()

webscout/Provider/OPENAI/README.md

@@ -21,9 +21,9 @@
 
 The WebScout OpenAI-Compatible Providers module offers a standardized way to interact with various AI providers using the familiar OpenAI API structure. This makes it easy to:
 
-*   Use the same code structure across different AI providers
-*   Switch between providers without major code changes
-*   Leverage the OpenAI ecosystem of tools and libraries with alternative AI providers
+- Use the same code structure across different AI providers
+- Switch between providers without major code changes
+- Leverage the OpenAI ecosystem of tools and libraries with alternative AI providers
 
 ## ⚙️ Available Providers
 
@@ -70,8 +70,9 @@ Currently, the following providers are implemented with OpenAI-compatible interf
 - FalconH1
 - XenAI
 - GeminiProxy
----
-
+- MonoChat
+- Friendli
+- MiniMax
 
 ## 💻 Usage Examples
 
@@ -909,17 +910,17 @@ All providers return responses that mimic the OpenAI API structure, ensuring com
 
 The OpenAI-compatible providers are built on a modular architecture:
 
-*   `base.py`: Contains abstract base classes that define the OpenAI-compatible interface
-*   `utils.py`: Provides data structures that mimic OpenAI's response format
-*   Provider-specific implementations (e.g., `deepinfra.py`): Implement the abstract interfaces for specific providers
+- `base.py`: Contains abstract base classes that define the OpenAI-compatible interface
+- `utils.py`: Provides data structures that mimic OpenAI's response format
+- Provider-specific implementations (e.g., `deepinfra.py`): Implement the abstract interfaces for specific providers
 
 This architecture makes it easy to add new providers while maintaining a consistent interface.
 
 ## 📝 Notes
 
-*   Some providers may require API keys for full functionality
-*   Not all OpenAI features are supported by all providers
-*   Response formats are standardized to match OpenAI's format, but the underlying content depends on the specific provider and model
+- Some providers may require API keys for full functionality
+- Not all OpenAI features are supported by all providers
+- Response formats are standardized to match OpenAI's format, but the underlying content depends on the specific provider and model
 
 ## 🤝 Contributing
 
@@ -932,24 +933,24 @@ Want to add a new OpenAI-compatible provider? Follow these steps:
 
 ## 📚 Related Documentation
 
-*   [OpenAI API Reference](https://platform.openai.com/docs/api-reference)
-*   [DeepInfra Documentation](https://deepinfra.com/docs)
-*   [Glider.so Website](https://glider.so/)
-*   [ChatGPT Clone Website](https://chatgpt-clone-ten-nu.vercel.app/)
-*   [X0GPT Website](https://x0-gpt.devwtf.in/)
-*   [WiseCat Website](https://wise-cat-groq.vercel.app/)
-*   [Venice AI Website](https://venice.ai/)
-*   [ExaAI Website](https://o3minichat.exa.ai/)
-*   [TypeGPT Website](https://chat.typegpt.net/)
-*   [SciraChat Website](https://scira.ai/)
-*   [FreeAIChat Website](https://freeaichatplayground.com/)
-*   [LLMChatCo Website](https://llmchat.co/)
-*   [Yep.com Website](https://yep.com/)
-*   [HeckAI Website](https://heck.ai/)
-*   [SonusAI Website](https://chat.sonus.ai/)
-*   [ExaChat Website](https://exa-chat.vercel.app/)
-*   [Netwrck Website](https://netwrck.com/)
-*   [StandardInput Website](https://chat.standard-input.com/)
+- [OpenAI API Reference](https://platform.openai.com/docs/api-reference)
+- [DeepInfra Documentation](https://deepinfra.com/docs)
+- [Glider.so Website](https://glider.so/)
+- [ChatGPT Clone Website](https://chatgpt-clone-ten-nu.vercel.app/)
+- [X0GPT Website](https://x0-gpt.devwtf.in/)
+- [WiseCat Website](https://wise-cat-groq.vercel.app/)
+- [Venice AI Website](https://venice.ai/)
+- [ExaAI Website](https://o3minichat.exa.ai/)
+- [TypeGPT Website](https://chat.typegpt.net/)
+- [SciraChat Website](https://scira.ai/)
+- [FreeAIChat Website](https://freeaichatplayground.com/)
+- [LLMChatCo Website](https://llmchat.co/)
+- [Yep.com Website](https://yep.com/)
+- [HeckAI Website](https://heck.ai/)
+- [SonusAI Website](https://chat.sonus.ai/)
+- [ExaChat Website](https://exa-chat.vercel.app/)
+- [Netwrck Website](https://netwrck.com/)
+- [StandardInput Website](https://chat.standard-input.com/)
 
 <div align="center">
   <a href="https://t.me/PyscoutAI"><img alt="Telegram Group" src="https://img.shields.io/badge/Telegram%20Group-2CA5E0?style=for-the-badge&logo=telegram&logoColor=white"></a>

webscout/Provider/OPENAI/TogetherAI.py

@@ -208,21 +208,15 @@ class TogetherAI(OpenAICompatibleProvider):
     OpenAI-compatible client for TogetherAI API.
     """
     AVAILABLE_MODELS = [
-        "Gryphe/MythoMax-L2-13b",
-        "Gryphe/MythoMax-L2-13b-Lite",
         "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
         "Qwen/QwQ-32B",
         "Qwen/Qwen2-72B-Instruct",
         "Qwen/Qwen2-VL-72B-Instruct",
         "Qwen/Qwen2.5-72B-Instruct-Turbo",
         "Qwen/Qwen2.5-7B-Instruct-Turbo",
-        "Qwen/Qwen2.5-Coder-32B-Instruct",
         "Qwen/Qwen2.5-VL-72B-Instruct",
-        "Qwen/Qwen3-235B-A22B-fp8",
         "Qwen/Qwen3-235B-A22B-fp8-tput",
-        "Rrrr/meta-llama/Llama-3-70b-chat-hf-6f9ad551",
-        "Rrrr/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo-03dc18e1",
-        "Rrrr/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo-6c92f39d",
+        "Salesforce/Llama-Rank-V1",
         "arcee-ai/arcee-blitz",
         "arcee-ai/caller",
         "arcee-ai/coder-large",
@@ -237,13 +231,12 @@ class TogetherAI(OpenAICompatibleProvider):
         "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
         "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B",
         "deepseek-ai/DeepSeek-V3",
-        "deepseek-ai/DeepSeek-V3-p-dp",
         "google/gemma-2-27b-it",
-        "google/gemma-2b-it",
         "lgai/exaone-3-5-32b-instruct",
         "lgai/exaone-deep-32b",
         "marin-community/marin-8b-instruct",
-        "meta-llama/Llama-3-70b-chat-hf",
+        "meta-llama-llama-2-70b-hf",
+        "meta-llama/Llama-2-70b-hf",
         "meta-llama/Llama-3-8b-chat-hf",
         "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo",
         "meta-llama/Llama-3.2-3B-Instruct-Turbo",
@@ -265,14 +258,8 @@ class TogetherAI(OpenAICompatibleProvider):
         "mistralai/Mixtral-8x7B-Instruct-v0.1",
         "nvidia/Llama-3.1-Nemotron-70B-Instruct-HF",
         "perplexity-ai/r1-1776",
-        "roberizk@gmail.com/meta-llama/Llama-3-70b-chat-hf-26ee936b",
-        "roberizk@gmail.com/meta-llama/Meta-Llama-3-70B-Instruct-6feb41f7",
-        "roberizk@gmail.com/meta-llama/Meta-Llama-3-8B-Instruct-8ced8839",
         "scb10x/scb10x-llama3-1-typhoon2-70b-instruct",
-        "scb10x/scb10x-llama3-1-typhoon2-8b-instruct",
-        "togethercomputer/MoA-1",
-        "togethercomputer/MoA-1-Turbo",
-        "togethercomputer/Refuel-Llm-V2",
+        "scb10x/scb10x-typhoon-2-1-gemma3-12b",
         "togethercomputer/Refuel-Llm-V2-Small",
     ]
 

webscout/Provider/OPENAI/__init__.py

@@ -43,7 +43,9 @@ from .PI import *  # Add PI.ai provider
 from .TogetherAI import *  # Add TogetherAI provider
 from .xenai import *  # Add XenAI provider
 from .GeminiProxy import *  # Add GeminiProxy provider
-
+from .friendli import *
+from .monochat import *
+from .MiniMax import *  # Add MiniMaxAI provider
 # Export auto-proxy functionality
 from .autoproxy import (
     get_auto_proxy,