Pricing

Choose the way you run OvalBee

Start with OvalBee Cloud for free, no credit card required, or run all of OvalBee on your own infrastructure with Enterprise.

Free

Go from raw images and videos to a trained model in one afternoon, with no credit card and nothing to install.

Free

 

Get started

10,000 credits/mo

Resets monthly. No top-ups.

Includes:

  • 2 seats, unlimited viewers
  • 10 GB of storage
  • Model training on GPUs
  • 100+ computer vision nodes
  • Ready-to-use workflows
  • Labeling, curation, and evaluation
  • AI agent

Pro

Bring your team, run bigger workloads, and take models to production with more credits and storage.

€80/mo

Billed annually

Choose Pro

Includes 100,000 credits/moEstimated task timeCPU~2,500 hoursfrom 40 credits/hourGPU~118 hoursfrom 850 credits/hourAllowance resets each billing periodTop-ups €10 / 10,000 credits · never expire

Everything in Free, plus:

  • 5 seats, more at €25/mo each
  • 100 GB of storage
  • Flow versioning and collaboration
  • Roles and permissions
  • Connect your own S3, Azure, or GCS bucket
  • Credit top-ups
  • Priority email support

Dedicated Cloud

Your own OvalBee instance, hosted and managed by us, with full code access and enterprise controls.

Custom

 

Talk to sales

Custom credits

Compute and storage sized to your team and workload.

Everything in Pro, plus:

  • Isolated instance hosted by OvalBee
  • Custom nodes with full code access
  • Single sign-on with OpenID Connect
  • Audit logs
  • Unlimited seats and storage
  • Custom contract, invoicing, and SLA
  • Expert onboarding and training

Full feature comparison

See how each plan compares across compute, labeling, curation, training, evaluation, integrations, and security.

Collaboration

Seats25 includedCustom
Extra seats—€25/mo eachCustom
ViewersUnlimitedUnlimitedUnlimited
Flow versioning—IncludedIncluded
Flow collaboration—IncludedIncluded
Roles and permissions—IncludedIncluded

Compute and storage

Included credits10,000/month100,000/monthCustom
Credit top-ups—€10 / 10,000Custom
CPU compute~250 h/month~2,500 h/monthCustom
GPU compute~12 h/month~118 h/monthCustom
Concurrent runs4 CPU · 1 GPU20 CPU · 4 GPUCustom
High-memory GPUs—IncludedIncluded
Included storage10 GB100 GBCustom
Hosting regionEUEUEU or custom

Labeling

Labeling and reviewIncludedIncludedIncluded
Consensus reviewIncludedIncludedIncluded
AI-assisted pre-labelingIncludedIncludedIncluded
Smart annotation toolsIncludedIncludedIncluded
Labeling queues and batchesIncludedIncludedIncluded

Data curation

Duplicate and near-duplicate searchIncludedIncludedIncluded
Outlier detectionIncludedIncludedIncluded
Label error detectionIncludedIncludedIncluded
Annotation gap detectionIncludedIncludedIncluded
Image quality checksIncludedIncludedIncluded
Semantic image searchIncludedIncludedIncluded
Dataset versioningIncludedIncludedIncluded
Annotation formatsIncludedIncludedIncluded

Training and deployment

Train modelsIncludedIncludedIncluded
Live training reportsIncludedIncludedIncluded
Model exportONNX / PyTorch / TensorRTONNX / PyTorch / TensorRTONNX / PyTorch / TensorRT
Deploy models as servicesIncludedIncludedIncluded
Running model deployments310Custom

Evaluation

Evaluation reportsIncludedIncludedIncluded
Model comparisonIncludedIncludedIncluded
Error analysisIncludedIncludedIncluded

AI agent

Build, edit, and fix flows from chatIncludedIncludedIncluded
MCP server for Claude Code, Cursor, and CodexIncludedIncludedIncluded

Flows and automation

FlowsUnlimitedUnlimitedUnlimited
Visual flow builderIncludedIncludedIncluded
100+ nodes includedIncludedIncludedIncluded
Ready-to-use workflowsIncludedIncludedIncluded
Conditional branchingIncludedIncludedIncluded
Flow import and exportIncludedIncludedIncluded
Webhook-triggered flowsIncludedIncludedIncluded
Scheduled nodesUnlimitedUnlimitedUnlimited
Custom nodes with full code access——Included

Integrations and connectivity

Python SDKIncludedIncludedIncluded
REST API and API tokensIncludedIncludedIncluded
S3-compatible APIIncludedIncludedIncluded
Email notificationsIncludedIncludedIncluded
Supervisely integrationIncludedIncludedIncluded
Own S3, Azure, or GCS buckets—1Custom

Support

DocumentationIncludedIncludedIncluded
Community forumIncludedIncludedIncluded
Priority email—IncludedIncluded
Expert onboarding and ML consulting——Included

Security and compliance

Encrypted secrets storeIncludedIncludedIncluded
Secure signed storageIncludedIncludedIncluded
Google and GitHub loginIncludedIncludedIncluded
SSO with OpenID Connect——Included
Audit logs——Included

Billing and contract

Service-level agreement——Included
Custom contract and invoicing——Included

FAQ

Pricing questions

Can I use OvalBee for free?

Yes. OvalBee Cloud has a free account with no credit card required, including a daily credit allowance to run workflows.

What is the difference between Cloud and Self-hosted?

Cloud is the hosted OvalBee service, with managed infrastructure and included compute credits. Self-hosted Enterprise runs on infrastructure you control, including air-gapped networks, and a free open-source Community edition is coming soon.

How do compute credits work?

Credits measure compute time, billed per second, and 1,000 credits cost €1. CPU tasks start at 40 credits per hour and GPU tasks at 850 credits per hour, with larger machines costing more. AI agent requests also use credits.

Can I add more credits without changing plans?

Yes, on Pro. Use the slider above to estimate the price of additional credits. Top-ups are not available on the Free plan, while Dedicated Cloud capacity is priced around your deployment.

Do unused credits roll over?

Included credits reset every month. Purchased top-up credits never expire and are used after the monthly allowance runs out.

How does annual billing work?

Cloud annual plans use the discounted rates shown on this page and are billed annually. Switch to Monthly above the plan cards whenever you want to compare the standard monthly rate.