feat: add MiniMax model support via OpenAI-compatible API

docs/examples/japanese_extraction.md

@@ -51,7 +51,7 @@ for entity in result.extractions:
     if entity.char_interval:
         start, end = entity.char_interval.start_pos, entity.char_interval.end_pos
         position_info = f" (pos: {start}-{end})"
-    
+
     print(f"• {entity.extraction_class}: {entity.extraction_text}{position_info}")
 
 # Expected Output:

langextract/_compat/exceptions.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Compatibility shim for langextract.exceptions imports."""
+
 # pylint: disable=duplicate-code
 
 from __future__ import annotations

langextract/_compat/registry.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Compatibility shim for langextract.registry imports."""
+
 # pylint: disable=duplicate-code
 
 from __future__ import annotations

langextract/_compat/schema.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Compatibility shim for langextract.schema imports."""
+
 # pylint: disable=duplicate-code
 
 from __future__ import annotations

langextract/core/base_model.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Base interfaces for language models."""
+
 from __future__ import annotations
 
 import abc

langextract/core/data.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Classes used to represent core data types of annotation pipeline."""
+
 from __future__ import annotations
 
 import dataclasses

langextract/core/debug_utils.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Debug utilities for LangExtract."""
+
 from __future__ import annotations
 
 import functools

langextract/core/schema.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Core schema abstractions for LangExtract."""
+
 from __future__ import annotations
 
 import abc

langextract/core/types.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Core data types for LangExtract."""
+
 from __future__ import annotations
 
 import dataclasses

langextract/data_lib.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Library for data conversion between AnnotatedDocument and JSON."""
+
 from __future__ import annotations
 
 import dataclasses

langextract/exceptions.py

@@ -17,6 +17,7 @@
 This module re-exports exceptions from core.exceptions for backward compatibility.
 All new code should import directly from langextract.core.exceptions.
 """
+
 # pylint: disable=duplicate-code
 
 from __future__ import annotations

langextract/io.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Supports Input and Output Operations for Data Annotations."""
+
 from __future__ import annotations
 
 import abc

langextract/plugins.py

@@ -17,6 +17,7 @@
 This module provides centralized provider discovery without circular imports.
 It supports both built-in providers and third-party providers via entry points.
 """
+
 from __future__ import annotations
 
 import functools

langextract/progress.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Progress and visualization utilities for LangExtract."""
+
 from __future__ import annotations
 
 from typing import Any

langextract/prompting.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Library for building prompts."""
+
 from __future__ import annotations
 
 import dataclasses

langextract/providers/builtin_registry.py

@@ -48,4 +48,9 @@ class ProviderConfig(TypedDict):
         'target': 'langextract.providers.openai:OpenAILanguageModel',
         'priority': patterns.OPENAI_PRIORITY,
     },
+    {
+        'patterns': patterns.MINIMAX_PATTERNS,
+        'target': 'langextract.providers.minimax:MiniMaxLanguageModel',
+        'priority': patterns.MINIMAX_PRIORITY,
+    },
 ]

langextract/providers/gemini.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """Gemini provider for LangExtract."""
+
 # pylint: disable=duplicate-code
 
 from __future__ import annotations

langextract/providers/minimax.py

@@ -0,0 +1,228 @@
+# Copyright 2025 Google LLC.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""MiniMax provider for LangExtract.
+
+This provider uses MiniMax's OpenAI-compatible API to extract structured
+information from text.
+
+Usage:
+    # Using factory
+    from langextract.factory import ModelConfig, create_model
+
+    config = ModelConfig(
+        model_id="MiniMax-M2.5",
+        provider="MiniMaxLanguageModel",
+        provider_kwargs={
+            "api_key": "your-minimax-api-key"
+        }
+    )
+    model = create_model(config)
+
+    result = lx.extract(
+        text_or_documents=text,
+        prompt_description=instructions,
+        model=model
+    )
+"""
+
+from __future__ import annotations
+
+import dataclasses
+from typing import Any
+
+from langextract.core import base_model
+from langextract.core import data
+from langextract.providers import patterns
+from langextract.providers import router
+
+_DEFAULT_MODEL_ID = "MiniMax-M2.5"
+_DEFAULT_BASE_URL = "https://api.minimax.io/v1"
+
+
+@router.register(
+    *patterns.MINIMAX_PATTERNS,
+    priority=patterns.MINIMAX_PRIORITY,
+)
+@dataclasses.dataclass(init=False)
+class MiniMaxLanguageModel(base_model.BaseLanguageModel):
+  """Language model inference using MiniMax's OpenAI-compatible API."""
+
+  model_id: str = _DEFAULT_MODEL_ID
+  api_key: str | None = None
+  base_url: str = _DEFAULT_BASE_URL
+  organization: str | None = None
+  format_type: data.FormatType = data.FormatType.JSON
+  temperature: float | None = None
+  max_workers: int = 10
+  _client: Any = dataclasses.field(default=None, repr=False, compare=False)
+  _extra_kwargs: dict[str, Any] = dataclasses.field(
+      default_factory=dict, repr=False, compare=False
+  )
+
+  @property
+  def requires_fence_output(self) -> bool:
+    """MiniMax returns raw JSON without fences."""
+    if self.format_type == data.FormatType.JSON:
+      return False
+    return super().requires_fence_output
+
+  def __post_init__(self):
+    """Initialize the OpenAI client with MiniMax configuration."""
+    try:
+      from openai import AsyncOpenAI
+    except ImportError as e:
+      raise ImportError(
+          "OpenAI package is required for MiniMax provider. "
+          "Install with: pip install langextract[openai]"
+      ) from e
+
+    if self._client is None:
+      self._client = AsyncOpenAI(
+          api_key=self.api_key,
+          base_url=self.base_url,
+          organization=self.organization,
+          **self._extra_kwargs,
+      )
+
+  async def _generate(
+      self,
+      texts: list[str],
+      prompt_description: str,
+      extra_params: dict[str, Any] | None = None,
+  ) -> list[list[base_model.ExtractionCandidate]]:
+    """Generate extractions for the given texts."""
+    import asyncio
+
+    extra_params = extra_params or {}
+
+    async def process_single(text: str) -> list[base_model.ExtractionCandidate]:
+      response = await self._client.chat.completions.create(
+          model=self.model_id,
+          messages=[
+              {
+                  "role": "system",
+                  "content": (
+                      "You are a helpful assistant that extracts structured"
+                      " information from text."
+                  ),
+              },
+              {
+                  "role": "user",
+                  "content": f"{prompt_description}\n\nText: {text}",
+              },
+          ],
+          response_format={"type": "json_object"}
+          if self.format_type == data.FormatType.JSON
+          else None,
+          temperature=self.temperature,
+          **extra_params,
+      )
+
+      content = response.choices[0].message.content
+      if not content:
+        return []
+
+      try:
+        import json
+
+        data = json.loads(content)
+        # Wrap in ExtractionCandidate format
+        if isinstance(data, list):
+          return [
+              base_model.ExtractionCandidate(
+                  extraction_text=item.get("text", str(item)),
+                  extraction_class=item.get("class", "unknown"),
+                  extraction_index=i,
+              )
+              for i, item in enumerate(data)
+          ]
+        elif isinstance(data, dict):
+          # For single object extractions
+          return [
+              base_model.ExtractionCandidate(
+                  extraction_text=str(v),
+                  extraction_class=k,
+                  extraction_index=i,
+              )
+              for i, (k, v) in enumerate(data.items())
+          ]
+      except (json.JSONDecodeError, AttributeError):
+        pass
+
+      return [
+          base_model.ExtractionCandidate(
+              extraction_text=content,
+              extraction_class="extracted",
+              extraction_index=0,
+          )
+      ]
+
+    # Process texts in parallel
+    tasks = [process_single(text) for text in texts]
+    results = await asyncio.gather(*tasks)
+    return results
+
+  def _generate_sync(
+      self,
+      texts: list[str],
+      prompt_description: str,
+      extra_params: dict[str, Any] | None = None,
+  ) -> list[list[base_model.ExtractionCandidate]]:
+    """Synchronous wrapper for generation."""
+    import asyncio
+
+    try:
+      loop = asyncio.get_event_loop()
+      if loop.is_running():
+        # If we're in an async context, we need to create a new loop
+        # This is a simplified sync wrapper - for production use async directly
+        import concurrent.futures
+
+        def run_in_executor():
+          return asyncio.run(
+              self._generate(texts, prompt_description, extra_params)
+          )
+
+        with concurrent.futures.ThreadPoolExecutor(max_workers=1) as executor:
+          future = executor.submit(run_in_executor)
+          return future.result()
+    except RuntimeError:
+      # No event loop, run directly
+      return asyncio.run(
+          self._generate(texts, prompt_description, extra_params)
+      )
+
+  def __call__(
+      self,
+      texts: Sequence[str],
+      prompt_description: str,
+      extra_params: dict[str, Any] | None = None,
+  ) -> list[list[base_model.ExtractionCandidate]]:
+    """Synchronous interface for the model."""
+    return self._generate_sync(list(texts), prompt_description, extra_params)
+
+  async def _call_async(
+      self,
+      texts: Sequence[str],
+      prompt_description: str,
+      extra_params: dict[str, Any] | None = None,
+  ) -> list[list[base_model.ExtractionCandidate]]:
+    """Asynchronous interface for the model."""
+    return await self._generate(list(texts), prompt_description, extra_params)
+
+  def close(self):
+    """Close the client connection."""
+    # AsyncOpenAI doesn't need explicit close
+    pass

langextract/providers/ollama.py

@@ -79,6 +79,7 @@
     2. Pull the model: ollama pull gemma2:2b
     3. Ollama server will start automatically when you use extract()
 """
+
 # pylint: disable=duplicate-code
 
 from __future__ import annotations

langextract/providers/openai.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 """OpenAI provider for LangExtract."""
+
 # pylint: disable=duplicate-code
 
 from __future__ import annotations