Marketplace · 85+ models
Computer Vision See what your business sees
Object detection, image classification, segmentation, OCR, facial recognition, video analytics, and pose estimation models — production-ready on Yobitel GPU Cloud.
Built by Yobitel · On AWS Marketplace
Yobitel CVAT AMI
Production-grade data annotation, ready in minutes
A pre-configured Amazon Machine Image of CVAT v1.11.0 on Ubuntu 24.04. Days of DevOps disappear — subscribe on AWS Marketplace, launch an EC2 instance, and your annotation team is labelling within minutes. Data stays inside your own AWS account; no per-seat licensing.
AI-assisted labelling out of the box
SAM cuts polygon time up to 70%. DEXTR draws boundaries from 4 clicks. SORT + SiamMask track objects across video frames automatically.
Every annotation type
Bounding boxes, polygons, polylines, points, cuboids, masks, keypoints, skeletons — for image and video datasets.
Training-ready exports
Direct export to YOLO, COCO, and Pascal VOC. No post-processing scripts. Goes from labels to training in one step.
Native S3, your account
IAM-role authentication, direct read/write to your S3 buckets. Data never leaves your AWS environment.
Extensible by design
Built on open-source CVAT — modify label schemas, drop in proprietary models via Nuclio, customize the Docker Compose stack.
Pay only for compute
Flat per-hour AWS Marketplace pricing scales with usage, not with team size or task volume.
At a glance
Stack & sizing
- Base OS
- Ubuntu 24.04
- CVAT version
- v1.11.0
- Stack
- Docker · PostgreSQL · Redis · Nginx
- Recommended start
- t3.large
- For AI-assist workloads
- m5.xlarge / m5.2xlarge
- Supported instance types
- 16 EC2 families
Why teams self-host
Compliance-grade by default
Healthcare
HIPAA
Clinical imaging data can't sit on third-party SaaS — CVAT AMI keeps PHI inside your AWS account.
Finance
GDPR
Annotating financial documents and PII under EU/regional data protection without exporting to a vendor.
Defence & Gov
Data sovereignty
Mission and intelligence imagery stays inside controlled infrastructure — no SaaS egress.
Enterprise
Internal InfoSec
When InfoSec blocks third-party data processing, self-host CVAT on your own EC2 in minutes.
Editor's pick
The computer vision model most teams reach for first.
YOLOv8
Real-time object detection across 5 model sizes; benchmark leader on COCO.
Spec sheet
- Family
- Ultralytics
- Parameters
- 3.2M – 68M
- License
- AGPL-3.0
- Status
- Live
- Best for
- See what your business sees
- Sits in
- Computer Vision
Pricing and routing rank visible on InferenceBench. Variants and quantisations appear in the Yobibyte deploy console.
The rest of the lineup
5 more in Computer Vision. All deployable in one click.
Promptable image and video segmentation with zero-shot generalization.
Transformer-based object detection without anchors or NMS post-processing.
Image-text dual encoder for zero-shot classification and retrieval.
Unified vision-language model: captioning, detection, segmentation, OCR.
Conditional image generation guided by edge, depth, pose, or scribble maps.
Showing 6 of 85+. The full catalog (with quantisations, hardware variants, and per-region pricing) lives in the Yobibyte console.
Quick start
Five lines to your first computer vision call.
Every model in this category is reachable from the same Yobitel SDK. Swap the model name; the rest of the call shape stays identical. Authenticated via your workspace key.
from yobitel import Inference
# YOLOv8 — real-time object detection
client = Inference(model="ultralytics/yolov8")
detections = client.detect(
image="warehouse_camera.jpg",
confidence_threshold=0.5,
)
for d in detections:
print(f"{d.label:12s} {d.confidence:.2f} box={d.box}")Where teams ship this
Real computer vision. In production.
Four use cases that customers run today. Pick a model from the lineup above, deploy on Yobibyte, plug it into the surrounding stack. Done.
- 01
Visual inspection in manufacturing
- 02
Retail analytics and shelf monitoring
- 03
Medical imaging triage
- 04
Autonomous systems and robotics
Frameworks
Bring what your team already knows
Yobitel handles the serving layer (GPU scheduling, KV cache, autoscaling, request batching) so your team focuses on the model and the product.
Learn about YobibyteExplore the rest
Other categories in the marketplace
NLP & Language
Text generation, translation, sentiment, summarization
Generative AI
Image gen, text gen, code gen, multimodal
Data Analytics
Predictive analytics, forecasting, anomaly detection
Automation & RPA
Process automation, workflow AI, document processing
Industry-Specific
Vertical-specific models by industry
Speech & Audio
ASR, TTS, speaker diarization, audio classification
Recommendation
Recommender systems, personalization, content matching
From the Yobitel blog
Deeper reads on Computer Vision
Yobitel.com
CVAT on AWS — how Yobitel's pre-built AMI makes AI data annotation effortless at scale
The full deployment story: what's pre-installed, AI-assist with SAM and DEXTR, S3 integration, and 16 supported instance types.
Read on yobitel.comYobitel.com
Yobitel CVAT Image & Video Annotation Solutions
How Yobitel CVAT handles image and video datasets end-to-end — from raw frames to training-ready exports.
Read on yobitel.comYobitel.com
Yobitel Multi-Model Text-to-Image Inference Server
The companion serving stack — multiple generative vision models behind one inference endpoint, GPU-optimised.
Read on yobitel.comDon't see what you need?
Bring your own model or fine-tune one of ours. Yobitel engineers can sit with your team and ship the right stack.