Pricing

Start free. Add support. Scale with the full platform.

One platform in every edition. Paid editions are priced on the environments, clusters and cloud accounts they cover, never on tokens or agent runtime.

  • Open source
  • Cloud SaaS or self-hosted
  • Air-gap ready
  • SOC 2 Type II
  • ISO 27001
Editions

Pick who runs it

Community

You deploy it. You run it.

FreeOpen source · unlimited clusters
Install with Helm
  • Every Task Bee in the catalog
  • All five Persona Bees
  • Convert runbooks; build your own
  • Typed actions and the trust dial
  • Four checks on every step
  • Full audit trail and kill switch
  • Bring your own model
  • Live service map of a cluster
  • VPC, on-prem or air-gapped
  • Community Slack and GitHub

Enterprise Support

You run it. We’re on call for you.

Starting at$500/ monthBilled annually · by environments and clusters
Get a support quote

Everything in Community, plus

  • bCortexMemory and a live knowledge graph, so every investigation starts from what your estate already knows
  • DAIM RouterThe right model for every step: better accuracy and latency at a lower cost
  • Help desk and support line
  • Critical issues in 1 hour, 24×7
  • Named support engineers
  • Priority fixes as hotfix builds
  • New model versions first
  • SSO, SCIM and audit-log export
  • 12-month LTS releases
  • Install and upgrade guidance
  • Annual health check
Full platform license

Enterprise Platinum

We implement it with you, in our cloud or yours.

Cloud SaaS or self-hostedContact usPriced by environments, clusters and accounts
Talk to us

Everything in Enterprise Support, plus

  • bCortexMemory and the knowledge graph across every cluster and account, with custom retention
  • DAIM RouterThe right model for every step, with a say in which new models are certified next

License

  • Commercial enterprise license
  • Cloud SaaS or self-hosted
  • 99.9% uptime SLA on SaaS
  • Region choice, custom retention

Implementation

  • Full install and implementation
  • Integrations set up with your stack
  • First runbooks converted to Task Bees
  • Trust-dial rollout plan

Product

  • Your requests on the roadmap
  • Priority releases and early access
  • Newest models first
  • 18-month LTS releases

Success

  • Named technical account manager
  • Critical issues in 30 min, 24×7
  • Shared Slack with our engineers
  • Quarterly reviews and training
  • Security review support

Every edition

  • The same Task Bees
  • The same guardrails
  • A full audit trail
  • Bring your own model
  • Self-hosted or air-gapped
Compare all features
FeatureCommunityEnterprise SupportEnterprise Platinum
Task Bees, Persona Bees, System BeesIncludedIncludedIncluded
Guardrails and the audit trailIncludedIncludedIncluded
DAIM RouterYou pick one model per step typeThe right model for every stepPlus a say in what is certified next
bCortex memory and knowledge graphLive map of a clusterEvery cluster and accountEvery cluster, custom retention
DeploymentSelf-hostedSelf-hostedCloud SaaS or self-hosted
LicenseOpen sourceOpen source + supportCommercial enterprise license
SupportCommunityHelp desk, 24×5; P1 24×7Named TAM, 24×7
Critical responseNot included1 hour30 minutes
SSO, SCIM, audit-log exportNot includedIncludedIncluded
ImplementationDocsGuidanceFull implementation
Roadmap priorityGitHub issuesConsideredPrioritized, early access
Uptime SLANot includedNot included99.9% on Cloud SaaS
AI models

Pick your model path

NudgeBee is model-agnostic. Choose how you pay for tokens, where models run and how you meet compliance, on any edition.

Bring your own model

Connect your LLM. Pay the vendor directly.

Zero NudgeBee markup

Best for
Existing LLM contracts, strict data residency or air-gapped environments
Billing
Pay as you go through your provider; no tokens routed through NudgeBee

NudgeBee managed model

Gemma and Qwen, tuned for cloud ops.

Flat annual fee, no token metering

Best for
Predictable AI cost without a separate LLM vendor
Billing
A flat annual fee for the models; infrastructure billed separately, your cloud or ours

Custom AI workload

Dedicated hosting and fine-tuned models.

Built around your compliance

Best for
Regulated industries, high throughput, fine-tuned or dedicated hosting
Billing
A custom quote covering models, throughput, hosting region and compliance

Running the managed model

Both options install Gemma and Qwen, tuned for cloud ops, inside your VPC.

OptionHost it yourselfWe host it for you
PricingYour GPU cost onlyA flat management fee plus your cloud
DeploymentAWS, Azure, GCP, on-prem or air-gappedFully managed by NudgeBee
FeesNo license fees, no per-token chargesNo per-token markup
UsageUnlimited within your hosting capacityUnlimited within your capacity

Supported models

  • Anthropic Claude
  • Google Gemini
  • OpenAI GPT
  • Self-hosted open-weightGemma, Llama, Nemotron, Qwen, DeepSeek

Your data is never used to train AI models. Not ours. Not anyone else’s.

Pricing FAQ

Questions Ops Teams Actually Ask

For your security team, procurement, and SRE lead.

Talk to Sales
We price on the environments you connect, such as production and staging, and the Kubernetes clusters and cloud accounts inside them. For clusters and accounts we use whichever count is greater, so 5 cloud accounts and 10 clusters count as 10.
The open-source platform: every Task Bee, all five Persona Bees, the guardrails and the full audit trail, on unlimited clusters. You deploy and run it yourself, with support from the community on Slack and GitHub.
bCortex and the DAIM Router, a help desk and support line with named engineers, critical issues answered within 1 hour, 24×7, hotfix builds, SSO and SCIM, and 12-month LTS releases. It starts at $500 a month, billed annually, and scales with the environments and clusters you cover.
When you want a commercial license and want us to do the work: Cloud SaaS or self-hosted, full installation and implementation, your requests prioritized on the roadmap, a named technical account manager and critical issues answered within 30 minutes, 24×7.
No. Paid editions are priced on environments, clusters and accounts, never on tokens or agent runtime. Bring your own model and you pay your provider directly, with no NudgeBee markup; the NudgeBee managed model is a flat annual fee.
Yes. Every edition runs the same platform, so your Task Bees, configuration and history carry over when you upgrade.
Yes. NudgeBee is available on AWS Marketplace, so enterprise procurement can purchase against existing AWS commits. Azure and GCP marketplace listings are on the roadmap.

Cloud Ops on Rails
With Human-in-the-Loop Guardrails.

One agentic platform for SRE, FinOps, and Kubernetes. Cut MTTR, reduce spend, and automate ops.

Book a Demo Start a Free Trial

Free to self-host  |  No credit card required

SOC2 Type 2 ISO 27001 Deploys in VPC BYOM