langchain-mixpeek

AI agents are blind. Mixpeek gives them eyes, ears, and memory.

The official LangChain integration for Mixpeek — the infrastructure layer that lets AI agents search video, images, audio, and documents through one API.

Quick install

pip install langchain-mixpeek

What's inside

Usage

Retriever

from langchain_mixpeek import MixpeekRetriever

retriever = MixpeekRetriever(
    api_key="mxp_...",
    retriever_id="ret_abc123",
    namespace="my-namespace",
)
docs = retriever.invoke("find the red cup")

One-line agent tool

tool = retriever.as_tool()  # that's it — retriever becomes an agent tool

Agent with 6 capabilities

from langchain_mixpeek import MixpeekToolkit
from langgraph.prebuilt import create_react_agent
from langchain_anthropic import ChatAnthropic

toolkit = MixpeekToolkit(
    api_key="mxp_...",
    namespace="brand-protection",
    bucket_id="bkt_...",
    collection_id="col_...",
    retriever_id="ret_...",
)

agent = create_react_agent(
    ChatAnthropic(model="claude-sonnet-4-20250514"),
    toolkit.get_tools(),
)

result = agent.invoke({
    "messages": [("user", "Scan these product URLs and alert me about counterfeits")]
})

The toolkit gives your agent:

Scope what your agent can do:

toolkit.get_tools(actions=["search"])                         # search only
toolkit.get_tools(actions=["search", "ingest", "process"])    # search + upload
toolkit.get_tools()                                           # all 6 tools

VectorStore — ingest any file type

from langchain_mixpeek import MixpeekVectorStore

store = MixpeekVectorStore(
    api_key="mxp_...",
    namespace="my-namespace",
    bucket_id="bkt_...",
    collection_id="col_...",
    retriever_id="ret_...",
)

# Eyes: images and video
store.add_images(["https://example.com/photo.jpg"])
store.add_videos(["https://example.com/clip.mp4"])

# Ears: audio
store.add_audio(["https://example.com/recording.mp3"])

# Memory: text, PDFs, spreadsheets
store.add_texts(["product description..."])
store.add_pdfs(["https://example.com/doc.pdf"])
store.add_excel(["https://example.com/data.xlsx"])

# Process everything (embedding, OCR, face detection, transcription)
store.trigger_processing()

# Search across all modalities
docs = store.similarity_search("red cup on the table")

Convert between interfaces anytime:

retriever = store.as_retriever()
tool = store.as_tool()
toolkit = store.as_toolkit()

RAG chain

from langchain_core.prompts import ChatPromptTemplate
from langchain_anthropic import ChatAnthropic
from langchain_mixpeek import MixpeekRetriever

retriever = MixpeekRetriever(api_key="mxp_...", retriever_id="ret_...", namespace="ns")
llm = ChatAnthropic(model="claude-sonnet-4-20250514")

prompt = ChatPromptTemplate.from_template(
    "Answer using this context:\n{context}\n\nQuestion: {question}"
)

chain = {"context": retriever, "question": lambda x: x} | prompt | llm
response = chain.invoke("what happens at 2 minutes?")

Multi-retriever agent

from langchain_mixpeek import MixpeekTool
from langgraph.prebuilt import create_react_agent

video_search = MixpeekTool(
    api_key="mxp_...", retriever_id="ret_video", namespace="archive",
    name="search_videos", description="Search video archive for scenes or faces.",
)
image_search = MixpeekTool(
    api_key="mxp_...", retriever_id="ret_images", namespace="catalog",
    name="search_images", description="Search product images by visual similarity.",
)

agent = create_react_agent(llm, [video_search, image_search])

Configuration

*Required for ingest/processing. Not needed for search-only via from_retriever().

Platform features

The VectorStore and Toolkit expose the full Mixpeek platform beyond search:

• Taxonomies — create and run classification pipelines on your documents
• Clusters — group similar content with kmeans, dbscan, hdbscan, or spectral clustering
• Alerts — monitor for matches and notify via webhook, Slack, or email
• Plugins — manage and test custom feature extractors

Documentation

• Full docs — tutorials, examples, platform features
• API reference — Python SDK reference
• Examples — runnable demo scripts

License

MIT