langchain - 💡(How to fix) Fix `embed_documents` returns 1 vector for N texts with `gemini-embedding-2`

Submission checklist

• This is a bug, not a usage question.
• I added a clear and descriptive title that summarizes this issue.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).
• This is not related to the langchain-community package.
• I posted a self-contained, minimal, reproducible example. A maintainer can copy it and run it AS IS.

Package (Required)

• langchain
• langchain-openai
• langchain-anthropic
• langchain-classic
• langchain-core
• langchain-model-profiles
• langchain-tests
• langchain-text-splitters
• langchain-chroma
• langchain-deepseek
• langchain-exa
• langchain-fireworks
• langchain-groq
• langchain-huggingface
• langchain-mistralai
• langchain-nomic
• langchain-ollama
• langchain-openrouter
• langchain-perplexity
• langchain-qdrant
• langchain-xai
• Other / not sure / general

Related Issues / PRs

No response

Reproduction Steps / Example Code (Python)

from langchain_google_genai import GoogleGenerativeAIEmbeddings

embeddings = GoogleGenerativeAIEmbeddings(
    model="models/gemini-embedding-2",
    output_dimensionality=768,
)
texts = ["hello world", "foo bar", "lorem ipsum"]
result = embeddings.embed_documents(texts)

print(len(result))   # prints 1  ← expected 3
print(len(result[0])) # prints 768

Error Message and Stack Trace (if applicable)

Description

GoogleGenerativeAIEmbeddings.embed_documents(["t1", "t2", ..., "tN"]) always returns a list of 1 vector regardless of how many texts are passed. embed_query("t1") works correctly and returns 1 vector for 1 text.

Root cause (investigated)

embed_documents calls batchEmbedContents with all texts as contents. gemini-embedding-2 treats a list of contents as parts of one multimodal document (by design for cross-modal retrieval), returning 1 combined vector instead of 1 vector per text.

Workaround

Embed each text individually via aembed_query using asyncio.gather:

async def embed_texts(texts): tasks = [embeddings.aembed_query(t) for t in texts] return list(await asyncio.gather(*tasks))

System Info

OS: Windows OS Version: 10.0.26200 Python Version: 3.14.4 (tags/v3.14.4:23116f9, Apr 7 2026, 14:10:54) [MSC v.1944 64 bit (AMD64)]

langchain_core: 1.4.0 langchain: 1.3.1 langchain_community: 0.4.2 langsmith: 0.8.5 langchain_classic: 1.0.7 langchain_google_genai: 4.2.3 langchain_protocol: 0.0.15 langchain_pymupdf4llm: 0.5.0 langchain_qdrant: 1.1.0 langchain_text_splitters: 1.1.2 langgraph_sdk: 0.3.15