langchain-google-genai 2.1.11__tar.gz → 3.0.0a1__tar.gz

{langchain_google_genai-2.1.11 → langchain_google_genai-3.0.0a1}/PKG-INFO

@@ -1,14 +1,16 @@
 Metadata-Version: 2.1
 Name: langchain-google-genai
-Version: 2.1.11
+Version: 3.0.0a1
 Summary: An integration package connecting Google's genai package and LangChain
 License: MIT
 Project-URL: Source Code, https://github.com/langchain-ai/langchain-google/tree/main/libs/genai
-Requires-Python: >=3.9
-Requires-Dist: langchain-core>=0.3.75
-Requires-Dist: google-ai-generativelanguage<1,>=0.7
-Requires-Dist: pydantic<3,>=2
-Requires-Dist: filetype<2,>=1.2
+Project-URL: Release Notes, https://github.com/langchain-ai/langchain-google/releases
+Project-URL: repository, https://github.com/langchain-ai/langchain-google
+Requires-Python: <4.0.0,>=3.10.0
+Requires-Dist: langchain-core<2.0.0,>=1.0.0a4
+Requires-Dist: google-ai-generativelanguage<1.0.0,>=0.7.0
+Requires-Dist: pydantic<3.0.0,>=2.0.0
+Requires-Dist: filetype<2.0.0,>=1.2.0
 Description-Content-Type: text/markdown
 
 # langchain-google-genai
@@ -54,6 +56,9 @@ This package provides LangChain support for Google Gemini models (via the offici
 
 ```bash
 pip install -U langchain-google-genai
+
+# or, with uv:
+uv add langchain-google-genai
 ````
 
 ---
@@ -71,7 +76,7 @@ Then use the `ChatGoogleGenerativeAI` interface:
 ```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="gemini-pro")
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash")
 response = llm.invoke("Sing a ballad of LangChain.")
 print(response.content)
 ```
@@ -80,22 +85,30 @@ print(response.content)
 
 ## Chat Models
 
-The main interface for Gemini chat models is `ChatGoogleGenerativeAI`.
+See the LangChain documentation for general information about [Chat Models](https://docs.langchain.com/oss/python/langchain/models).
+
+The main interface for the Gemini chat models is `ChatGoogleGenerativeAI`.
 
 ### Multimodal Inputs
 
-Gemini vision models support image inputs in single messages.
+Most Gemini models support image inputs.
 
 ```python
 from langchain_core.messages import HumanMessage
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="gemini-pro-vision")
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash")
 
 message = HumanMessage(
     content=[
-        {"type": "text", "text": "What's in this image?"},
-        {"type": "image_url", "image_url": "https://picsum.photos/seed/picsum/200/300"},
+        {
+            "type": "text",
+            "text": "What's in this image?"
+        },
+        {
+            "type": "image_url",
+            "image_url": "https://picsum.photos/seed/picsum/200/300"
+        },
     ]
 )
 
@@ -103,7 +116,7 @@ response = llm.invoke([message])
 print(response.content)
 ```
 
-✅ `image_url` can be:
+`image_url` can be:
 
 - A public image URL
 - A Google Cloud Storage path (`gcs://...`)
@@ -113,38 +126,44 @@ print(response.content)
 
 ### Multimodal Outputs
 
-The Gemini 2.0 Flash Experimental model supports both text and inline image outputs.
+Some Gemini models supports both text and inline image outputs.
 
 ```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="models/gemini-2.0-flash-exp-image-generation")
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash-image-preview")
 
 response = llm.invoke(
     "Generate an image of a cat and say meow",
     generation_config=dict(response_modalities=["TEXT", "IMAGE"]),
 )
 
-image_base64 = response.content[0].get("image_url").get("url").split(",")[-1]
-meow_text = response.content[1]
+image_base64 = response.content[1].get("image_url").get("url").split(",")[-1]
+meow_text = response.content[0]
 print(meow_text)
+# In Jupyter, display the image:
+from base64 import b64decode
+from IPython.display import Image, display
+
+img_bytes = b64decode(image_base64)
+display(Image(data=img_bytes))
 ```
 
 ---
 
 ### Audio Output
 
-```
+```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="models/gemini-2.5-flash-preview-tts")
-# example
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash-preview-tts")
+
 response = llm.invoke(
     "Please say The quick brown fox jumps over the lazy dog",
     generation_config=dict(response_modalities=["AUDIO"]),
 )
 
-# Base64 encoded binary data of the image
+# Base64 encoded binary data of the audio
 wav_data = response.additional_kwargs.get("audio")
 with open("output.wav", "wb") as f:
     f.write(wav_data)
@@ -154,15 +173,13 @@ with open("output.wav", "wb") as f:
 
 ### Multimodal Outputs in Chains
 
-You can use Gemini models in a LangChain chain:
-
 ```python
 from langchain_core.runnables import RunnablePassthrough
 from langchain_core.prompts import ChatPromptTemplate
 from langchain_google_genai import ChatGoogleGenerativeAI, Modality
 
 llm = ChatGoogleGenerativeAI(
-    model="models/gemini-2.0-flash-exp-image-generation",
+    model="gemini-2.5-flash-image-preview",
     response_modalities=[Modality.TEXT, Modality.IMAGE],
 )
 
@@ -178,13 +195,11 @@ response = chain.invoke("cat")
 
 ### Thinking Support
 
-Gemini 2.5 Flash Preview supports internal reasoning ("thoughts").
-
 ```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
 llm = ChatGoogleGenerativeAI(
-    model="models/gemini-2.5-flash-preview-04-17",
+    model="models/gemini-2.5-flash",
     thinking_budget=1024
 )
 
@@ -199,8 +214,6 @@ print("Reasoning tokens used:", reasoning_score)
 
 ## Embeddings
 
-You can use Gemini embeddings in LangChain:
-
 ```python
 from langchain_google_genai import GoogleGenerativeAIEmbeddings
 
@@ -249,4 +262,4 @@ print("Answerable probability:", response.answerable_probability)
 
 - [LangChain Documentation](https://docs.langchain.com/)
 - [Google Generative AI SDK](https://googleapis.github.io/python-genai/)
-- [Gemini Model Documentation](https://ai.google.dev/)
+- [Gemini Model Documentation](https://ai.google.dev/gemini-api/docs)

{langchain_google_genai-2.1.11 → langchain_google_genai-3.0.0a1}/README.md

@@ -41,6 +41,9 @@ This package provides LangChain support for Google Gemini models (via the offici
 
 ```bash
 pip install -U langchain-google-genai
+
+# or, with uv:
+uv add langchain-google-genai
 ````
 
 ---
@@ -58,7 +61,7 @@ Then use the `ChatGoogleGenerativeAI` interface:
 ```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="gemini-pro")
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash")
 response = llm.invoke("Sing a ballad of LangChain.")
 print(response.content)
 ```
@@ -67,22 +70,30 @@ print(response.content)
 
 ## Chat Models
 
-The main interface for Gemini chat models is `ChatGoogleGenerativeAI`.
+See the LangChain documentation for general information about [Chat Models](https://docs.langchain.com/oss/python/langchain/models).
+
+The main interface for the Gemini chat models is `ChatGoogleGenerativeAI`.
 
 ### Multimodal Inputs
 
-Gemini vision models support image inputs in single messages.
+Most Gemini models support image inputs.
 
 ```python
 from langchain_core.messages import HumanMessage
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="gemini-pro-vision")
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash")
 
 message = HumanMessage(
     content=[
-        {"type": "text", "text": "What's in this image?"},
-        {"type": "image_url", "image_url": "https://picsum.photos/seed/picsum/200/300"},
+        {
+            "type": "text",
+            "text": "What's in this image?"
+        },
+        {
+            "type": "image_url",
+            "image_url": "https://picsum.photos/seed/picsum/200/300"
+        },
     ]
 )
 
@@ -90,7 +101,7 @@ response = llm.invoke([message])
 print(response.content)
 ```
 
-✅ `image_url` can be:
+`image_url` can be:
 
 - A public image URL
 - A Google Cloud Storage path (`gcs://...`)
@@ -100,38 +111,44 @@ print(response.content)
 
 ### Multimodal Outputs
 
-The Gemini 2.0 Flash Experimental model supports both text and inline image outputs.
+Some Gemini models supports both text and inline image outputs.
 
 ```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="models/gemini-2.0-flash-exp-image-generation")
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash-image-preview")
 
 response = llm.invoke(
     "Generate an image of a cat and say meow",
     generation_config=dict(response_modalities=["TEXT", "IMAGE"]),
 )
 
-image_base64 = response.content[0].get("image_url").get("url").split(",")[-1]
-meow_text = response.content[1]
+image_base64 = response.content[1].get("image_url").get("url").split(",")[-1]
+meow_text = response.content[0]
 print(meow_text)
+# In Jupyter, display the image:
+from base64 import b64decode
+from IPython.display import Image, display
+
+img_bytes = b64decode(image_base64)
+display(Image(data=img_bytes))
 ```
 
 ---
 
 ### Audio Output
 
-```
+```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
-llm = ChatGoogleGenerativeAI(model="models/gemini-2.5-flash-preview-tts")
-# example
+llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash-preview-tts")
+
 response = llm.invoke(
     "Please say The quick brown fox jumps over the lazy dog",
     generation_config=dict(response_modalities=["AUDIO"]),
 )
 
-# Base64 encoded binary data of the image
+# Base64 encoded binary data of the audio
 wav_data = response.additional_kwargs.get("audio")
 with open("output.wav", "wb") as f:
     f.write(wav_data)
@@ -141,15 +158,13 @@ with open("output.wav", "wb") as f:
 
 ### Multimodal Outputs in Chains
 
-You can use Gemini models in a LangChain chain:
-
 ```python
 from langchain_core.runnables import RunnablePassthrough
 from langchain_core.prompts import ChatPromptTemplate
 from langchain_google_genai import ChatGoogleGenerativeAI, Modality
 
 llm = ChatGoogleGenerativeAI(
-    model="models/gemini-2.0-flash-exp-image-generation",
+    model="gemini-2.5-flash-image-preview",
     response_modalities=[Modality.TEXT, Modality.IMAGE],
 )
 
@@ -165,13 +180,11 @@ response = chain.invoke("cat")
 
 ### Thinking Support
 
-Gemini 2.5 Flash Preview supports internal reasoning ("thoughts").
-
 ```python
 from langchain_google_genai import ChatGoogleGenerativeAI
 
 llm = ChatGoogleGenerativeAI(
-    model="models/gemini-2.5-flash-preview-04-17",
+    model="models/gemini-2.5-flash",
     thinking_budget=1024
 )
 
@@ -186,8 +199,6 @@ print("Reasoning tokens used:", reasoning_score)
 
 ## Embeddings
 
-You can use Gemini embeddings in LangChain:
-
 ```python
 from langchain_google_genai import GoogleGenerativeAIEmbeddings
 
@@ -236,4 +247,4 @@ print("Answerable probability:", response.answerable_probability)
 
 - [LangChain Documentation](https://docs.langchain.com/)
 - [Google Generative AI SDK](https://googleapis.github.io/python-genai/)
-- [Gemini Model Documentation](https://ai.google.dev/)
+- [Gemini Model Documentation](https://ai.google.dev/gemini-api/docs)

{langchain_google_genai-2.1.11 → langchain_google_genai-3.0.0a1}/langchain_google_genai/__init__.py

@@ -1,4 +1,4 @@
-"""**LangChain Google Generative AI Integration**
+"""**LangChain Google Generative AI Integration**.
 
 This module integrates Google's Generative AI models, specifically the Gemini series, with the LangChain framework. It provides classes for interacting with chat models and generating embeddings, leveraging Google's advanced AI capabilities.
 
@@ -76,12 +76,12 @@ __all__ = [
     "AqaOutput",
     "ChatGoogleGenerativeAI",
     "DoesNotExistsException",
+    "DoesNotExistsException",
     "GenAIAqa",
-    "GoogleGenerativeAIEmbeddings",
     "GoogleGenerativeAI",
+    "GoogleGenerativeAIEmbeddings",
     "GoogleVectorStore",
     "HarmBlockThreshold",
     "HarmCategory",
     "Modality",
-    "DoesNotExistsException",
 ]

{langchain_google_genai-2.1.11 → langchain_google_genai-3.0.0a1}/langchain_google_genai/_common.py

@@ -13,19 +13,17 @@ _TELEMETRY_ENV_VARIABLE_NAME = "GOOGLE_CLOUD_AGENT_ENGINE_ID"
 
 
 class GoogleGenerativeAIError(Exception):
-    """
-    Custom exception class for errors associated with the `Google GenAI` API.
-    """
+    """Custom exception class for errors associated with the `Google GenAI` API."""
 
 
 class _BaseGoogleGenerativeAI(BaseModel):
-    """Base class for Google Generative AI LLMs"""
+    """Base class for Google Generative AI LLMs."""
 
     model: str = Field(
         ...,
         description="""The name of the model to use.
 Examples:
-    - gemini-2.5-pro
+    - gemini-2.5-flash
     - models/text-bison-001""",
     )
     """Model name to use."""
@@ -34,28 +32,38 @@ Examples:
     )
     """Google AI API key.
     If not specified will be read from env var ``GOOGLE_API_KEY``."""
+
     credentials: Any = None
     "The default custom credentials (google.auth.credentials.Credentials) to use "
     "when making API calls. If not provided, credentials will be ascertained from "
     "the GOOGLE_API_KEY envvar"
+
     temperature: float = 0.7
-    """Run inference with this temperature. Must be within ``[0.0, 2.0]``."""
+    """Run inference with this temperature. Must be within ``[0.0, 2.0]``. If unset,
+    will default to ``0.7``."""
+
     top_p: Optional[float] = None
     """Decode using nucleus sampling: consider the smallest set of tokens whose
-       probability sum is at least ``top_p``. Must be within ``[0.0, 1.0]``."""
+    probability sum is at least ``top_p``. Must be within ``[0.0, 1.0]``."""
+
     top_k: Optional[int] = None
     """Decode using top-k sampling: consider the set of ``top_k`` most probable tokens.
-       Must be positive."""
+    Must be positive."""
+
     max_output_tokens: Optional[int] = Field(default=None, alias="max_tokens")
     """Maximum number of tokens to include in a candidate. Must be greater than zero.
-       If unset, will default to ``64``."""
+    If unset, will use the model's default value, which varies by model.
+    See https://ai.google.dev/gemini-api/docs/models for model-specific limits."""
+
     n: int = 1
     """Number of chat completions to generate for each prompt. Note that the API may
-       not return the full ``n`` completions if duplicates are generated."""
-    max_retries: int = 6
-    """The maximum number of retries to make when generating."""
+    not return the full ``n`` completions if duplicates are generated."""
+
+    max_retries: int = Field(default=6, alias="retries")
+    """The maximum number of retries to make when generating. If unset, will default
+    to ``6``."""
 
-    timeout: Optional[float] = None
+    timeout: Optional[float] = Field(default=None, alias="request_timeout")
     """The maximum number of seconds to wait for a response."""
 
     client_options: Optional[Dict] = Field(
@@ -68,6 +76,7 @@ Examples:
     transport: Optional[str] = Field(
         default=None,
         description="A string, one of: [`rest`, `grpc`, `grpc_asyncio`].",
+        alias="api_transport",
     )
     additional_headers: Optional[Dict[str, str]] = Field(
         default=None,
@@ -89,9 +98,9 @@ Examples:
     )
 
     safety_settings: Optional[Dict[HarmCategory, HarmBlockThreshold]] = None
-    """The default safety settings to use for all generations. 
-    
-        For example: 
+    """The default safety settings to use for all generations.
+
+        For example:
 
         .. code-block:: python
             from google.generativeai.types.safety_types import HarmBlockThreshold, HarmCategory
@@ -127,6 +136,7 @@ def get_user_agent(module: Optional[str] = None) -> Tuple[str, str]:
     Args:
         module (Optional[str]):
             Optional. The module for a custom user agent header.
+
     Returns:
         Tuple[str, str]
     """
@@ -148,11 +158,13 @@ def get_client_info(module: Optional[str] = None) -> "ClientInfo":
     Args:
         module (Optional[str]):
             Optional. The module for a custom user agent header.
+
     Returns:
         ``google.api_core.gapic_v1.client_info.ClientInfo``
     """
     client_library_version, user_agent = get_user_agent(module)
-    return ClientInfo(
+    # TODO: remove ignore once google-auth has types.
+    return ClientInfo(  # type: ignore[no-untyped-call]
         client_library_version=client_library_version,
         user_agent=user_agent,
     )