Find any scene in your video library.

Describe what you are looking for in plain English and get the exact matches back, across video, images, audio, and documents. One multimodal API on the object storage you already use.

Bring vectors

MVS

Agent-native vector store on object storage. Dense, sparse, and BM25 search. First 1M vectors free forever.

Start with MVS Read MVS docs

Connect files

Managed

Managed indexing extracts scenes, faces, OCR, transcripts, and embeddings from any file type.

VideoImagesAudioPDFsText

Start with Managed Talk to us

Case Study

A performance-video agency maps the creative DNA of 170K+ scenes across 8 brands: search any moment, cluster what works, reuse the winners. See how

Built for production retrieval

1Mvectors free, no expiration

$0.023per GB-month of vector storage

$1per million queries

Billionsvectors, scale on object storage

Perception layer

What the agent sees.

Every object you index becomes structured, searchable features: faces and objects in a frame, layout regions in a document, speakers in audio. Those are the same features an agent queries, and joins across modalities.

Hover or tap a card to preview the search it powers.

Person walking past a storefront, with face, body, and handbag detected

Person · 0.95Face · 0.98Handbag · 0.91Video · 00:04:12

scene caption

“A woman in a white shirt walks past a red storefront.”

transcript · “…meet me at the corner in five.”

A printed resume with header, body, chart, and signature regions detected

HeaderChartBodySignaturePDF · resume.pdf

layout extraction

Header, body, charts, and signature detected as typed regions.

OCR · 1 header · 3 sections · 1 signature

A studio microphone, with audio diarized into speaker segments

Speaker A

Speaker B

Speaker A

match · “limited time offer”

speaker diarization

Who spoke when, aligned to the transcript and the timeline.

audio · 2 speakers · matched at 00:01:30

View the extractors that produce these features

Cross-modal joins

One query across every modality.

Real questions rarely fit one feature. “Find the moment our CEO said guidance while the slide read Q4 outlook” needs a face, a spoken phrase, and on-screen text to line up at the same instant.

Mixpeek ties those features to the same object and timestamp, so an agent gets back the exact clip instead of three unrelated matches.

See how multi-stage retrieval joins features

One query, three features

Video

face_id = person_42

Audio

transcript contains “guidance”

Document

on-screen text reads “Q4 outlook”

⨝ join on object + timestamp

earnings_call.mp4 · 00:12:040.93

The CEO says “guidance” as the slide behind her reads “Q4 outlook.”

matched onface_idtranscripton-screen text

Quickstart

One install. Two paths.

Most retrieval stacks mean gluing together a vector DB, a file pipeline, and an agent layer. Mixpeek is one install with two ways in.

Install

pip install mixpeek

MVS quickstart

Bring embeddings

Create MVS namespace Read MVS docs

# pip install mixpeek
from mixpeek import Mixpeek

mp = Mixpeek(api_key="YOUR_KEY")

mp.namespaces.documents.upsert(
    namespace="agent_memory",
    documents=[{
        "id": "clip_001",
        "dense_embedding": embedding,
        "metadata": {"source": "s3://media/clip_001.mp4"}
    }]
)

results = mp.namespaces.documents.search(
    namespace="agent_memory",
    queries=[{"vector": query_embedding, "top_k": 10}]
)

01BYO embeddings

02Object storage

03Agent retrieval

Integrations

Plugs into your existing stack.

Connect your storage, point Mixpeek at it, and every file becomes searchable by what's inside it. No migration, no code changes.

storage

Mux

Every Mux upload becomes searchable by face, scene, transcript, and on-screen text, with no manual tagging.

View integration →

storage

Backblaze B2

S3-compatible extraction at 1/5th the cost. Store on B2, extract with Mixpeek, zero egress fees.

View integration →

storage

Iconik

Every asset in your DAM becomes findable by what's inside it: scenes, faces, spoken words, on-screen text.

View integration →

View all integrations →

Real workflows

In production right now.

National Gallery · live demo

Visual search across 45k artworks

Upload any image and find visually similar paintings across 45,000+ artworks, or just describe what you're looking for. Hybrid image and text retrieval, ranked with RRF.

Try gallery search →

Movie taste · live demo

Posters that learn your taste

Like or dislike movie posters and watch the grid adapt to your taste in real time. Interaction signals feed learned fusion so recommendations improve from usage.

Try movie personalization →

Super Bowl corpus · live demo

Face search across video

Drop in a headshot and find every clip a person appears in across 63 video ads and 2,600+ faces. Full trace for takedown evidence.

Try face search →

View others →

Extractors

Turn any file into searchable features.

Connect a bucket and these pipelines run as they are. Every one is documented and open source in the extractor cookbook.

Multimodal (Video/Audio/Image)

Video · Image · Audio · Text

Unified embeddings for video, audio, image, and text. Scene and silence chunking, Whisper transcription, thumbnails.

Try it live Docs GitHub

Universal All-in-One

Any file

One extractor for image, video, audio, and documents. Auto-detects modality and applies the right pipeline.

Try it live Docs GitHub

Image Embeddings (SigLIP)

Image · PDF

Dense 768-D image embeddings with Google SigLIP for text-to-image search in one contrastive space.

Try it live Docs GitHub

Text Embeddings (E5-Large)

Text

Multilingual dense text embeddings with E5-Large for semantic search and RAG out of the box.

Try it live Docs GitHub

Multi-File Object Embeddings (Gemini)

Any file

Embed ALL files of an object (images, PDFs, video, audio, text) into one 3072-D Gemini vector.

Try it live Docs GitHub

Face Identity (SCRFD + ArcFace)

Image · Video · PDF

Production face recognition that detects, aligns, and embeds faces to 512-D ArcFace vectors.

Try it live Docs GitHub

View all extractors

Enterprise-ready

SOC 2 Type II HIPAA-ready BYO-Cloud Self-hosted option Audit trails SSO / SAML

Pricing

Free vectors. Usage-based indexing.

Two ways to pay. MVS is priced by the vector and starts with 1M free. Managed is priced by the object, with credits covering extraction, embedding, and indexing.

MVS

1Mvectors free

Bring your own embeddings and pay by the vector. Dense, sparse, and BM25 search on your own object storage, with no expiration on the free tier.

Start with MVS

Managed

$0.001/credit

Bring raw objects and pay by the object. Credits cover extraction, embedding, indexing, enrichment, and retrieval.

Start with Managed

Enterprise

Custom

Dedicated infrastructure, self-hosted options, SSO, SLA, security reviews, and hands-on architecture support.

Talk to us

See full pricing

FAQ

Common questions.

Do I have to move my data?

No. Mixpeek reads from your existing S3, GCS, R2, Azure, or S3-compatible bucket. Your storage stays the system of record, and nothing leaves your cloud.

How fast is retrieval?

Hybrid queries (dense, sparse, and BM25) return in well under 100ms p95, even with vectors persisted on object storage rather than held in RAM.

Do I need embeddings to start?

No. Bring your own vectors with MVS, or point Managed at raw files and it generates embeddings and features for you.

What can Managed extract?

Faces, scenes, transcripts, OCR, labels, and embeddings from video, images, audio, PDFs, and documents, all indexed at the object level.

Can I self-host?

Yes. Deploy in your own cloud (BYO-Cloud) with SOC 2 and HIPAA-ready controls, SSO, audit trails, and namespaces.

How does pricing work?

MVS starts free with 1M vectors and no expiration. Managed is usage-based credits covering extraction, embedding, indexing, and retriever execution.