feat: add MiniMax as a first-class LLM and TTS provider

.env-template

@@ -8,6 +8,9 @@ INTERNAL_KEY=<internal key for worker-to-backend authentication>
 EMBEDDINGS_BASE_URL=
 EMBEDDINGS_KEY=
 
+#For MiniMax (optional - for using MiniMax models)
+MINIMAX_API_KEY=
+
 #For Azure (you can delete it if you don't use Azure)
 OPENAI_API_BASE=
 OPENAI_API_VERSION=

README.md

@@ -41,7 +41,7 @@
     <li><strong>🔑 Streamlined API Keys:</strong>  Generate keys linked to your settings, documents, and models, simplifying chatbot and integration setup.</li>
     <li><strong>🔗 Actionable Tooling:</strong> Connect to APIs, tools, and other services to enable LLM actions.</li>
     <li><strong>🧩 Pre-built Integrations:</strong> Use readily available HTML/React chat widgets, search tools, Discord/Telegram bots, and more.</li>
-    <li><strong>🔌 Flexible Deployment:</strong> Works with major LLMs (OpenAI, Google, Anthropic) and local models (Ollama, llama_cpp).</li>
+    <li><strong>🔌 Flexible Deployment:</strong> Works with major LLMs (OpenAI, Google, Anthropic, MiniMax) and local models (Ollama, llama_cpp).</li>
     <li><strong>🏢 Secure & Scalable:</strong> Run privately and securely with Kubernetes support, designed for enterprise-grade reliability.</li>
 </ul>
 

application/core/model_configs.py

@@ -193,6 +193,34 @@
     ),
 ]
 
+MINIMAX_ATTACHMENTS = IMAGE_ATTACHMENTS
+
+MINIMAX_MODELS = [
+    AvailableModel(
+        id="MiniMax-M2.5",
+        provider=ModelProvider.MINIMAX,
+        display_name="MiniMax M2.5",
+        description="MiniMax flagship model with 204K context window",
+        capabilities=ModelCapabilities(
+            supports_tools=True,
+            supported_attachment_types=MINIMAX_ATTACHMENTS,
+            context_window=204800,
+        ),
+    ),
+    AvailableModel(
+        id="MiniMax-M2.5-highspeed",
+        provider=ModelProvider.MINIMAX,
+        display_name="MiniMax M2.5 Highspeed",
+        description="Fast, cost-effective MiniMax model with 204K context window",
+        capabilities=ModelCapabilities(
+            supports_tools=True,
+            supported_attachment_types=MINIMAX_ATTACHMENTS,
+            context_window=204800,
+        ),
+    ),
+]
+
+
 AZURE_OPENAI_MODELS = [
     AvailableModel(
         id="azure-gpt-4",

application/core/model_settings.py

@@ -19,6 +19,7 @@ class ModelProvider(str, Enum):
     PREMAI = "premai"
     SAGEMAKER = "sagemaker"
     NOVITA = "novita"
+    MINIMAX = "minimax"
 
 
 @dataclass
@@ -118,6 +119,10 @@ def _load_models(self):
             settings.LLM_PROVIDER == "huggingface" and settings.API_KEY
         ):
             self._add_huggingface_models(settings)
+        if settings.MINIMAX_API_KEY or (
+            settings.LLM_PROVIDER == "minimax" and settings.API_KEY
+        ):
+            self._add_minimax_models(settings)
         # Default model selection
         if settings.LLM_NAME:
             # Parse LLM_NAME (may be comma-separated)
@@ -273,6 +278,21 @@ def _add_huggingface_models(self, settings):
         )
         self.models[model_id] = model
 
+    def _add_minimax_models(self, settings):
+        from application.core.model_configs import MINIMAX_MODELS
+
+        if settings.MINIMAX_API_KEY:
+            for model in MINIMAX_MODELS:
+                self.models[model.id] = model
+            return
+        if settings.LLM_PROVIDER == "minimax" and settings.LLM_NAME:
+            for model in MINIMAX_MODELS:
+                if model.id == settings.LLM_NAME:
+                    self.models[model.id] = model
+                    return
+        for model in MINIMAX_MODELS:
+            self.models[model.id] = model
+
     def _parse_model_names(self, llm_name: str) -> List[str]:
         """
         Parse LLM_NAME which may contain comma-separated model names.

application/core/model_utils.py

@@ -14,6 +14,7 @@ def get_api_key_for_provider(provider: str) -> Optional[str]:
         "google": settings.GOOGLE_API_KEY,
         "groq": settings.GROQ_API_KEY,
         "huggingface": settings.HUGGINGFACE_API_KEY,
+        "minimax": settings.MINIMAX_API_KEY,
         "azure_openai": settings.API_KEY,
         "docsgpt": None,
         "llama.cpp": None,

application/core/settings.py

@@ -90,6 +90,7 @@ class Settings(BaseSettings):
     GROQ_API_KEY: Optional[str] = None
     HUGGINGFACE_API_KEY: Optional[str] = None
     OPEN_ROUTER_API_KEY: Optional[str] = None
+    MINIMAX_API_KEY: Optional[str] = None
 
     OPENAI_API_BASE: Optional[str] = None  # azure openai api base url
     OPENAI_API_VERSION: Optional[str] = None  # azure openai api version
@@ -174,6 +175,7 @@ class Settings(BaseSettings):
         "GOOGLE_API_KEY",
         "GROQ_API_KEY",
         "HUGGINGFACE_API_KEY",
+        "MINIMAX_API_KEY",
         "EMBEDDINGS_KEY",
         "FALLBACK_LLM_API_KEY",
         "QDRANT_API_KEY",

application/llm/llm_creator.py

@@ -9,6 +9,7 @@
 from application.llm.openai import AzureOpenAILLM, OpenAILLM
 from application.llm.premai import PremAILLM
 from application.llm.sagemaker import SagemakerAPILLM
+from application.llm.minimax import MiniMaxLLM
 from application.llm.open_router import OpenRouterLLM
 
 logger = logging.getLogger(__name__)
@@ -27,6 +28,7 @@ class LLMCreator:
         "google": GoogleLLM,
         "novita": NovitaLLM,
         "openrouter": OpenRouterLLM,
+        "minimax": MiniMaxLLM,
     }
 
     @classmethod

application/llm/minimax.py

@@ -0,0 +1,74 @@
+from application.core.settings import settings
+from application.llm.openai import OpenAILLM
+
+MINIMAX_BASE_URL = "https://api.minimax.io/v1"
+
+
+class MiniMaxLLM(OpenAILLM):
+    def __init__(self, api_key=None, user_api_key=None, base_url=None, *args, **kwargs):
+        super().__init__(
+            api_key=api_key or settings.MINIMAX_API_KEY or settings.API_KEY,
+            user_api_key=user_api_key,
+            base_url=base_url or MINIMAX_BASE_URL,
+            *args,
+            **kwargs,
+        )
+
+    @staticmethod
+    def _clamp_temperature(kwargs):
+        """Clamp temperature to MiniMax's valid range (0.0, 1.0]."""
+        if "temperature" in kwargs:
+            temp = kwargs["temperature"]
+            if temp is not None:
+                temp = float(temp)
+                if temp <= 0:
+                    kwargs["temperature"] = 0.01
+                elif temp > 1.0:
+                    kwargs["temperature"] = 1.0
+
+    def _raw_gen(
+        self,
+        baseself,
+        model,
+        messages,
+        stream=False,
+        tools=None,
+        response_format=None,
+        **kwargs,
+    ):
+        self._clamp_temperature(kwargs)
+        # MiniMax does not support response_format; drop it
+        return super()._raw_gen(
+            baseself,
+            model,
+            messages,
+            stream=stream,
+            tools=tools,
+            response_format=None,
+            **kwargs,
+        )
+
+    def _raw_gen_stream(
+        self,
+        baseself,
+        model,
+        messages,
+        stream=True,
+        tools=None,
+        response_format=None,
+        **kwargs,
+    ):
+        self._clamp_temperature(kwargs)
+        # MiniMax does not support response_format; drop it
+        return super()._raw_gen_stream(
+            baseself,
+            model,
+            messages,
+            stream=stream,
+            tools=tools,
+            response_format=None,
+            **kwargs,
+        )
+
+    def _supports_structured_output(self):
+        return False

application/tts/minimax_tts.py

@@ -0,0 +1,74 @@
+import base64
+import json
+import logging
+
+import requests
+
+from application.core.settings import settings
+from application.tts.base import BaseTTS
+
+logger = logging.getLogger(__name__)
+
+MINIMAX_TTS_BASE_URL = "https://api.minimax.io"
+
+MINIMAX_TTS_VOICES = [
+    "English_Graceful_Lady",
+    "English_Insightful_Speaker",
+    "English_radiant_girl",
+    "English_Persuasive_Man",
+    "English_Lucky_Robot",
+    "English_expressive_narrator",
+]
+
+
+class MiniMaxTTS(BaseTTS):
+    def __init__(self):
+        self.api_key = settings.MINIMAX_API_KEY
+        self.base_url = MINIMAX_TTS_BASE_URL
+
+    def text_to_speech(self, text):
+        lang = "en"
+        url = f"{self.base_url}/v1/t2a_v2"
+
+        payload = {
+            "model": "speech-2.8-hd",
+            "text": text,
+            "stream": False,
+            "voice_setting": {
+                "voice_id": "English_Graceful_Lady",
+                "speed": 1,
+                "vol": 1,
+                "pitch": 0,
+            },
+            "audio_setting": {
+                "sample_rate": 32000,
+                "bitrate": 128000,
+                "format": "mp3",
+                "channel": 1,
+            },
+        }
+
+        response = requests.post(
+            url,
+            headers={
+                "Content-Type": "application/json",
+                "Authorization": f"Bearer {self.api_key}",
+            },
+            json=payload,
+            timeout=30,
+        )
+        response.raise_for_status()
+
+        result = response.json()
+        status_code = result.get("base_resp", {}).get("status_code", -1)
+        if status_code != 0:
+            status_msg = result.get("base_resp", {}).get("status_msg", "unknown error")
+            raise RuntimeError(f"MiniMax TTS error ({status_code}): {status_msg}")
+
+        hex_audio = result.get("data", {}).get("audio", "")
+        if not hex_audio:
+            raise RuntimeError("MiniMax TTS returned empty audio data")
+
+        audio_bytes = bytes.fromhex(hex_audio)
+        audio_base64 = base64.b64encode(audio_bytes).decode("utf-8")
+        return audio_base64, lang

application/tts/tts_creator.py

@@ -1,5 +1,6 @@
 from application.tts.google_tts import GoogleTTS
 from application.tts.elevenlabs import ElevenlabsTTS
+from application.tts.minimax_tts import MiniMaxTTS
 from application.tts.base import BaseTTS
 
 
@@ -8,6 +9,7 @@ class TTSCreator:
     tts_providers = {
         "google_tts": GoogleTTS,
         "elevenlabs": ElevenlabsTTS,
+        "minimax_tts": MiniMaxTTS,
     }
 
     @classmethod

docs/content/Models/cloud-providers.mdx

@@ -33,6 +33,7 @@ DocsGPT offers direct, streamlined support for the following cloud LLM providers
 | Prem AI                      | `premai`       | (See Prem AI docs)          |
 | AWS SageMaker                | `sagemaker`    | (See SageMaker docs)        |
 | Novita AI                    | `novita`       | (See Novita docs)           |
+| MiniMax                      | `minimax`      | `MiniMax-M2.5`              |
 
 ## Connecting to OpenAI-Compatible Cloud APIs
 

setup.ps1

@@ -268,9 +268,10 @@ function Prompt-CloudAPIProviderOptions {
     Write-ColorText "5) HuggingFace Inference API" -ForegroundColor "Yellow"
     Write-ColorText "6) Azure OpenAI" -ForegroundColor "Yellow"
     Write-ColorText "7) Novita" -ForegroundColor "Yellow"
+    Write-ColorText "8) MiniMax" -ForegroundColor "Yellow"
     Write-ColorText "b) Back to Main Menu" -ForegroundColor "Yellow"
     Write-Host ""
-    $script:provider_choice = Read-Host "Choose option (1-7, or b)"
+    $script:provider_choice = Read-Host "Choose option (1-8, or b)"
 }
 
 # Function to prompt for Ollama CPU/GPU options
@@ -982,11 +983,18 @@ function Connect-CloudAPIProvider {
                 Get-APIKey
                 break
             }
+            "8" {  # MiniMax
+                $script:provider_name = "MiniMax"
+                $script:llm_name = "minimax"
+                $script:model_name = "MiniMax-M2.5"
+                Get-APIKey
+                break
+            }
             "b" { Clear-Host; return }
             "B" { Clear-Host; return }
             default {
                 Write-Host ""
-                Write-ColorText "Invalid choice. Please choose 1-7, or b." -ForegroundColor "Red"
+                Write-ColorText "Invalid choice. Please choose 1-8, or b." -ForegroundColor "Red"
                 Start-Sleep -Seconds 1
             }
         }

setup.sh

@@ -154,9 +154,10 @@ prompt_cloud_api_provider_options() {
     echo -e "${YELLOW}5) HuggingFace Inference API${NC}"
     echo -e "${YELLOW}6) Azure OpenAI${NC}"
     echo -e "${YELLOW}7) Novita${NC}"
+    echo -e "${YELLOW}8) MiniMax${NC}"
     echo -e "${YELLOW}b) Back to Main Menu${NC}"
     echo
-    read -p "$(echo -e "${DEFAULT_FG}Choose option (1-7, or b): ${NC}")" provider_choice
+    read -p "$(echo -e "${DEFAULT_FG}Choose option (1-8, or b): ${NC}")" provider_choice
 }
 
 # Function to prompt for Ollama CPU/GPU options
@@ -707,8 +708,14 @@ connect_cloud_api_provider() {
                 model_name="deepseek/deepseek-r1"
                 get_api_key
                 break ;;
+            8) # MiniMax
+                provider_name="MiniMax"
+                llm_provider="minimax"
+                model_name="MiniMax-M2.5"
+                get_api_key
+                break ;;
             b|B) clear; return 1 ;; # Clear screen and Back to Main Menu
-            *) echo -e "\n${RED}Invalid choice. Please choose 1-7, or b.${NC}" ; sleep 1 ;;
+            *) echo -e "\n${RED}Invalid choice. Please choose 1-8, or b.${NC}" ; sleep 1 ;;
         esac
     done