MediaLayer

Media Similarity & Search Infrastructure

One platform. Two products.

Match API handles real-time one-to-one comparison across image, video, and audio. MediaLayer Search indexes your entire media library and supports low-latency search across large indexed collections.

Use Match API when you want to compare two media URLs. Use MediaLayer Search when you want to index a library and search one asset against many.

Match API

Real-time one-to-one comparison

Submit two media URLs and get a similarity score, matched segments, and confidence in a single low-latency request. No library to maintain.

  • Image, video, and audio similarity API
  • URL-based — no uploads required
  • Segment-level and near-duplicate matching
  • Same JSON envelope across all media types
  • RapidAPI subscriptions for self-serve access
  • Free plan to evaluate

MediaLayer Search

Indexed one-to-many media search

Index your entire media library — videos, audio, images — and run real-time fingerprint searches across millions of assets with low-latency results.

  • One query against millions of indexed assets
  • Image, audio, and video fingerprinting
  • Async ingestion for large workloads
  • Multi-tenant with full data isolation
  • Designed for low-latency search across large indexed libraries
  • Managed onboarding and infrastructure setup

Match API · 1-to-1

Developer-friendly. Pay as you grow.

Subscribe on RapidAPI and start comparing media immediately. Same endpoints, same JSON envelope across every tier. Upgrade or downgrade any time.

Testing & prototypes

Basic

$0/mo

Requests

100 / month

Rate limit

10 req / min

Production apps

Pro

$25/mo

Requests

5,000 / month

Rate limit

20 req / min

Most popular

Scale workloads

Ultra

$75/mo

Requests

20,000 / month

Rate limit

60 req / min

High throughput

Mega

$150/mo

Requests

50,000 / month

Rate limit

120 req / min

All plans include image, video, and audio similarity matching · Same API endpoints across every tier · Managed via RapidAPI

Infrastructure

What is Search Processing?

Search Processing measures the amount of media and fingerprint data processed during similarity search. Larger query files, longer videos, longer audio clips, and deeper candidate reranking consume more Search Processing than small image searches.

MediaLayer Search is priced by indexed media capacity and monthly Search Processing, not raw query count. This keeps pricing fair because one small image search and one long video search do not consume the same infrastructure.

🗂️

Fingerprint index search

Each query searches optimized fingerprint indexes to retrieve and rerank the most likely matches across your indexed media library.

Candidate reranking

Shortlisted candidates are precisely reranked using perceptual similarity scoring — Hamming distance for image and video, offset voting for audio.

🔄

Concurrent query handling

Multiple search requests are processed in parallel. Larger query files and longer media consume more search infrastructure per request.

📊

Shared pool across media types

Search Processing is a single shared pool across image, audio, and video. One allocation covers all active media types — no per-module search billing.

📡

Result delivery

Matched segments, similarity scores, and timestamps are returned per query. Video results carry more data than image or audio results.

📐

Approx. queries per tier

Shared pool — your media mix determines actual capacity

Tier🎬 Video🎵 Audio🖼️ Image
Starter~5K~50K~500K
Growth~20K~200K~2M
Scale~60K~600K~6M

Estimates based on approximate query media size / processing footprint, not API response size.

Indexed media capacity

Why capacity is split by media type

Images, audio, and video do not consume the same indexing resources. A single image, a one-hour audio file, and a thirty-minute video are very different infrastructure workloads. MediaLayer Search separates image count, audio hours, and video hours so pricing stays clear and predictable.

🖼️

Images — priced by count

Each image generates a fixed fingerprint at ingest. Images are flat-cost assets: 32 fingerprint rows per image, no duration. Indexed image capacity is measured in item count.

🎵

Audio — priced by hours

Audio indexing cost scales with duration. A 1-hour podcast generates far more fingerprint data than a 30-second clip. Indexed audio capacity is measured in total hours.

🎬

Video — priced by hours

Video is the most infrastructure-intensive media type. Each second generates 128 fingerprint rows across 32 perceptual hash bands. Indexed video capacity is measured in total hours.

Find your plan

Not sure which plan fits?

Estimate your workload below — we'll recommend the right plan based on your indexed media capacity and Search Processing needs.

Only need image, audio, or video search? Select just the media types you need — the estimator combines the pricing automatically.

Estimate your Search plan

Select which media types you need and adjust your workload — we'll combine the pricing and recommend the right plan.

What do you want to index?

Select one or more — prices are combined automatically.

Under 10K

≈ small product catalog or image archive

Under 250 GB

~5K video · ~50K audio · ~500K image queries / mo

Shared pool across all selected media types — charged once.

Recommended plan

Image Search — Starter

Estimated price

$19/mo

Managed setup · Provisioned within 24 hrs

Your image indexing and Search Processing requirements fit within the Starter plan.

Included

10K indexed images
250 GB/month Search Processing

Compare

Which product fits your workflow?

FeatureMatch APISearchEnterprise
Image similarity matching
Video similarity matching
Audio similarity matching
One-to-one comparison
One-to-many indexed search
Segment-level matching
Indexed image datasets
Indexed audio datasets
Indexed video datasets
Search Processing
Custom indexing pipelines
SLA supportScale planCustom
Dedicated infrastructure
Private / VPC deployment
On-prem deployment
Security review
Architecture reviewScale plan
Priority supportScale plan

Capacity

Need higher capacity?

MediaLayer Search plans are designed around bundled infrastructure capacity. If your workload exceeds the included image, audio, video, or Search Processing limits, you can upgrade to the next plan or request a custom Search plan.

Use cases

Built for media infrastructure workloads

MediaLayer is purpose-built signal processing infrastructure for teams that need reliable, fast, and accurate media similarity search at scale.

📺

Ad-tech creative audit

Detect recycled creatives, flag non-compliant edits, and verify campaign assets against approved masters — across image, video, and audio.

🛒

Marketplace deduplication

Identify copied product images and near-duplicate listings at ingestion time. Prevent fraud before it reaches search results.

©️

Copyright detection

Index reference libraries and surface unauthorized reuse — clipped segments, re-encodes, watermark removals — in real time.

🛡️

Content moderation

Build similarity-aware moderation queues. Match incoming content against known-bad media libraries at upload time, not after.

🎬

Enterprise media indexing

Index production-scale media libraries — broadcast archives, OTT catalogues, ad asset vaults — and run fingerprint search against them with low-latency retrieval.

🔊

Audio content monitoring

Detect re-uploaded music, sound recordings, and broadcast audio across platforms. Offset-aligned matching handles partial clips and re-encodes.

FAQ

Common questions

What is the difference between Match API and MediaLayer Search?
Match API (1-to-1) compares two media files on demand — stateless, per-request, no library to maintain. MediaLayer Search (1-to-N) indexes your entire media library and lets you query one clip against millions in real time. They serve different access patterns and are priced differently.
What is Search Processing?
Search Processing measures the amount of media and fingerprint data processed during similarity search. Larger query files, longer videos, longer audio clips, and deeper candidate reranking consume more Search Processing than small image searches. We price by Search Processing rather than raw query count because one small image search and one long video search do not consume the same infrastructure.
Do you offer image-only, audio-only, or video-only Search plans?
Yes. MediaLayer Search supports any combination of media types. Use the pricing estimator, select the media types you need, and it will calculate the combined price automatically. You only pay for the modules you activate.
Why are the main pricing cards shown as All Media plans?
The main Search plans show bundled image, audio, and video capacity for teams that need cross-media search. If you only need one or two media types, the pricing estimator will show a lower combined price for your specific combination.
Are single-media plans add-ons?
No. Selecting only image, audio, or video in the estimator gives you a standalone Search plan for that media type. Plans are not add-ons and do not stack on top of the bundled plans.
Why are images, audio, and video limited separately?
Each media type consumes different indexing resources. A single image generates a flat fingerprint at ingest. A one-hour audio file generates far more fingerprint data than a 30-second clip. Video is the most intensive — each second produces hundreds of fingerprint rows. Separating capacity by media type keeps pricing clear and predictable.
What happens if I exceed my included limits?
If your workload exceeds the included image, audio, video, or Search Processing limits, you can upgrade to the next plan or contact us for a custom Search plan. Public plans are bundled to keep pricing predictable — there are no self-serve overage charges.
How does indexed search work?
You ingest your media library via URL or bulk upload. Our pipeline fingerprints each asset and stores the fingerprints in an optimized index. At query time, the index is searched to retrieve and rerank the most likely matches — designed for low-latency results across large indexed libraries.
Does MediaLayer store my media files?
No. MediaLayer stores fingerprints only — compact mathematical representations derived from your media. Source files are downloaded during ingest and immediately discarded after fingerprinting. We never retain your original content.
Can MediaLayer run in a private or VPC deployment?
Yes, for Enterprise accounts. We support hosted (managed by us), VPC (your cloud, our stack), and on-prem deployments. Deployment preference is scoped during the onboarding conversation.
Can I get custom limits?
Yes. Enterprise plans are fully custom — indexed media capacity, Search Processing limits, infrastructure, and SLA are all scoped to your workload. Contact sales to start the conversation.
How does managed onboarding work?
Tell us your use case, expected library size, and media types. We provision your account, API keys, and infrastructure within 24 hours. Every account is set up by our team so infrastructure is sized correctly from day one.
What is the Match API rate limit?
Rate limits are set by your RapidAPI tier: 10 req/min on Basic, 20 on Pro, 60 on Ultra, 120 on Mega. For higher concurrency, contact us for direct API access outside of RapidAPI.

More questions? Contact MediaLayer AI Labs.

Start matching. We'll handle the setup.

Try Match API free on RapidAPI. For MediaLayer Search, tell us your use case and your account will be ready within 24 hours.

Need the API reference? Open the docs.