Event photo API · Endurance sports specialist

Race Photo API Built for Endurance — Not Generic Cloud Face Tools

Looking past Amazon Rekognition, Azure Face, or Google Cloud Vision for marathon and race galleries? 9Pic API delivers selfie search and bib / race-number search trained on endurance media — and handles the media backend, indexing, and storage so you do not rebuild an event stack around a generic vision endpoint.

Why not AWS / Azure / GCP vision APIs? · Capabilities · Docs & coding agents

AWS, Azure, and GCP vision APIs are generic — races are not

Amazon Rekognition, Azure Face, Google Cloud Vision, and similar cloud vision services are powerful building blocks for identity and labeling. Endurance photo delivery needs a full event media system: motion blur, partial bibs, helmets, multi-search, and searchable galleries at race volume.

What you still build on generic cloud vision APIs

  • Event model, photo ingest, indexing jobs, and retry queues
  • Storage layout, thumbnails, permissions, and download delivery
  • Bib / race-number OCR tuned for running and motorsport plates
  • Participant search UX (selfie + bib + name) and support for missed matches
  • False matches and OCR misses that feel like “hallucinations” on sweaty, angled race shots

What 9Pic API ships for endurance teams

  • FaceFind selfie search + BibTrack bib search in one API surface
  • Media backend for high-volume race libraries — not a single raw detect endpoint
  • Photo storage with you or with us (BYOS S3-compatible buckets)
  • Specialized for marathons, trail, triathlon, cycling, OCR, and race-number sports
  • API docs + AI-readable pages so coding agents integrate faster

Race & timing apps

Embed selfie and bib search in your registration or results app without standing up a computer-vision team.

White-label galleries

Keep your brand and UX. Call 9Pic for search, indexing, and media delivery behind the scenes.

Teams that already have AI

Prototyped on open-source face models, Rekognition, Azure Face, or Cloud Vision? Graduate to production race search without rewriting everything yourself.

Specialized for endurance sports photo search

Hyperscaler face and vision APIs optimize for clean portraits, documents, and general images. 9Pic optimizes for race-day reality across running, multi-sport, and mass participation events.

Motion & crowd shots

Finish chutes, pack running, and panning shots where faces and bibs are imperfect.

Partial bib match

Hands, belts, and folds hide digits — partial search still surfaces the right runner.

Selfie when bibs fail

Sunglasses, rain jackets, and blocked numbers: FaceFind recovers photos generic OCR alone misses.

Race numbers & plates

Alphanumeric codes for waves, categories, and motorsport-style identifiers — not only clean 1:1 face ID.

Explore verticals: marathon, triathlon, cycling, OCR, motorsports.

Core API Capabilities

Three powerful APIs — powered by the same AI engines behind FaceFind, BibTrack, and Motion.

Powered by 9Pic BibTrack

BIB Number Search API

Search across thousands of event photos by race number. Our BIB detection engine handles motion blur, partial visibility, angled shots, and crowded frames — returning accurate results even in the toughest conditions.

  • Motion blur resistant — built for high-speed sports
  • Partial match support for partially hidden BIBs
  • Works across marathons, cycling, triathlons, and more
POST /api/v1/search/bib
{
  "event_id": "marathon-2026",
  "bib_number": "4521",
  "limit": 20
}

// Response
{
  "results": [
    { "photo_id": "img_8a2f...", "confidence": 0.97 },
    { "photo_id": "img_3c1b...", "confidence": 0.94 }
  ]
}
POST /api/v1/search/face
{
  "event_id": "conference-2026",
  "selfie_url": "https://...",
  "threshold": 0.85,
  "limit": 50
}

// Response
{
  "results": [
    { "photo_id": "img_7d4e...", "similarity": 0.98 },
    { "photo_id": "img_2a9c...", "similarity": 0.96 }
  ]
}
Powered by 9Pic FaceFind

Selfie Search API

Let your users find their photos with a selfie. Upload a face image, and our AI matches it against every photo in the gallery — returning results in under a second with 99.7% accuracy.

  • 99.7% face recognition accuracy
  • Sub-second response times at scale
  • Works for events, conferences, weddings, and more
Powered by 9Pic Motion

Video Timestamp & Clipping API

Got a long video stream from a race or event? Our API pinpoints the exact timestamps where a specific person appears — using either their BIB number or a selfie. Get precise start and end times so you can generate participant-specific clips automatically.

  • Exact timestamp extraction from long video streams
  • Search by BIB number or selfie within video
  • Near-realtime clipping — results in under 10 seconds
POST /api/v1/video/clip
{
  "event_id": "marathon-2026",
  "video_id": "vid_finish_line",
  "bib_number": "4521"
}

// Response
{
  "clips": [
    { "start": "01:23:45.200", "end": "01:23:52.800" },
    { "start": "02:10:11.400", "end": "02:10:18.100" }
  ]
}

How 9Pic API works

We operate the search backend and media pipeline. You keep the customer-facing product.

1

Create access

Sign up, open Developer Zone, and create an API key. Start from the API reference.

2

Connect storage

Use 9Pic-managed photo storage or bring your own S3-compatible bucket. We handle indexing and search state.

3

Call face + bib APIs

REST endpoints for selfie search, bib search, listing photos, and video timestamps — documented for humans and coding agents.

4

Go live under your brand

Participants search in your app or site. You own UX; we run the endurance-grade search stack.

Docs and coding-agent friendly setup

Your developers should not reverse-engineer a generic vision SDK. Use our published docs — and feed them to AI coding tools so integration starts in minutes, not weeks of plumbing.

Use a coding agent skill on day one

Install official 9Pic Skills so Cursor, Claude Code, Codex, or your preferred coding agent scaffolds API calls against real endpoints — instead of inventing a half-finished face pipeline.

npx skills add https://docs.9pic.ai

Photo storage included — or bring your own

With generic face APIs you still own every byte of gallery storage, CDN, and lifecycle. 9Pic can host event media for you, or connect to your existing object storage while we run search and indexing.

9Pic-managed storage

We hold event media end-to-end

Google Cloud Storage

Your GCP bucket

Azure Blob Storage

Your Azure account

Amazon S3 & S3-compatible

Your object storage endpoint

Full media backend, not only embeddings

Ingest, process, search, and deliver race photos through one platform. Keep objects on Amazon S3, Google Cloud Storage, Azure Blob, or another S3-compatible endpoint if you want — without rebuilding gallery and search services around a raw vision call.

Also works with common S3-compatible providers (DigitalOcean Spaces, Backblaze B2, MinIO, Wasabi, and custom endpoints). Details in the API docs and onboarding with our team.

Ship your product with AI coding tools — without inventing race AI

Modern teams scaffold apps with coding agents in days. The hard part is still reliable selfie and bib search at race scale. 9Pic is that specialized backend.

Coding agent + 9Pic skill path

Run npx skills add https://docs.9pic.ai, then give the agent the Agent Skills docs and llms.txt so it wires auth, face search, and bib search correctly — instead of hallucinating a brittle face pipeline. See the install guide.

Production accuracy for race media

Specialized FaceFind and BibTrack engines reduce false matches and OCR misses that generic Rekognition, Azure Face, Cloud Vision, and document OCR leave for your team to paper over.

Scale for mass participation

From small club races to tens of thousands of runners — the same search stack used across high-volume endurance events.

Dashboard + API

Monitor processing and events in the 9Pic dashboard while your product calls the API under your brand.

Frequently Asked Questions

For endurance and mass-participation photo delivery, yes — because 9Pic is an event photo platform API, not a general computer-vision toolkit. Amazon Rekognition, Azure Face, Google Cloud Vision, and similar services are excellent generic building blocks, but you still assemble storage, event models, bib OCR, search UX, and race-day edge cases. 9Pic packages selfie search, bib search, media backend, and optional storage for that use case — whatever cloud you already use for your app.

Yes. Create an account, generate an API key in Developer Zone, and follow the API reference. For coding agents, run npx skills add https://docs.9pic.ai and point them at the Agent Skills docs and llms.txt. Enterprise volume and complex BYOS setups can still be scoped with our team.

Core endpoints include BIB number search (BibTrack), selfie/face search (FaceFind), video timestamp extraction and clipping (Motion), event photo listing, and downloads. See the published API docs for the full surface.

Yes. You can use 9Pic-managed event media storage, or bring your own object storage (including Google Cloud Storage, Azure Blob, Amazon S3, and other S3-compatible endpoints). Either way, 9Pic runs indexing and search so you are not only buying raw embeddings.

Yes. Processing status, event management, and operational visibility remain in the 9Pic AI dashboard while participants use your product.

Any stack that can call HTTPS REST APIs — mobile apps, web apps, timing platforms, custom galleries, and backend services.

The API exposes BibTrack, FaceFind, and Motion engines plus photo/event data access. Products such as Certify, Buzz, Flow, and Checkout are platform features when you use the full 9Pic experience; API customers typically implement their own front-end and commerce.

From the repo where the agent writes code, run npx skills add https://docs.9pic.ai. That installs official 9Pic Agent Skills for API auth, participant galleries, and partner-owned media sales. Details: Agent Skills docs and the install guide.

Ready to Get Started?

Join event organizers who've delivered 100M+ photos to 5M+ participants across 5k+ events using 9Pic AI for hosting, AI search, and media delivery.